Skip to main content

Full text of "Proceedings of the First Conference on Scientific Research in the National Parks"

See other formats

I* ## » 


Clemson University 


"\ ~ 

'Moo ■ 

3 1604 003 618 404 

J ) 




■ — ' 












/ . - -<■- ■ 








n V 







i "o 



J fe£ 

^ ^2 




^l U w 

tX fr: 








; ^ 



^ e^ 






^boo z 

MAY . 








Volume I 

Digitized by the Internet Archive 

in 2013 

Proceedings of the 

First Conference on Scientific Research 

in the National Parks 

Volume I 

New Orleans, Louisiana 
November 9-12,1 976 

Edited by Robert M. Linn 

Sponsored By: 

National Park Service and American Institute of Biological Sciences 

U.S. Department of the Interior 

National Park Service Transactions and Proceedings Series • Number Five • 1979 

As the Nation's principal conservation agency, the Department of the Interior has 
responsibility for most of our nationally owned public lands and natural resources. This 
includes fostering the wisest use of our land and water resources, protecting our fish and 
wildlife, preserving the environmental and cultural values of our national parks and 
historical places, and providing for the enjoyment of life through outdoor recreation. 
The Department assesses our energy and mineral resources and works to assure that 
their development is in the best interests of all our people. The Department also has a 
major responsibility for American Indian reservation communities and for people who 
live in Island Territories under U.S. administration. 

Note: This publication contains all Conference papers received by the editor's deadline. 

Library of Congress Cataloging In Publication Data 

Conference on Scientific Research in the National Parks, 

1st, New Orleans, 1976. 

Proceedings of the First Conference on Scientific 
Research in the National Parks, New Orleans, Louisiana, 
November 9-12, 1976. 

(Transactions and proceedings series - National Park 
Service; no. 5) 

Includes bibliographies. 

1. Natural history-United States-Congresses. 
2. Science-United States-Congresses. 3. National parks 
and reserves-United States-Congresses. 
I. Linn, Robert M. II. United States. National Park 
Service. III. American Institute of Biological Sciences. 
IV. Series: United States. National Park Service. 
Transactions and proceedings series - National Park 
Service; no. 5. 

QH104.C59 1976 500.9'73'072 78-21700 

For sale by the Superintendent of Documents, U.S. Government Printing Office 
Washington, D.C. 20402 

Stock No. 024-OO5-O0743-1 



Opening Remarks - ROBERT M. LINN xv 

Welcome to the Conference - RICHARD TRUMBULL xvii 

Introduction of Nathaniel P. Reed - THEODORE W. SUDIA xix 

Keynote Address - GARY EVERHARDT xxi 

Closing Remarks - ROBERT M. LINN xxvii 

Summary of Conference - STANLEY A. CAIN xxix 

TV Editorial xxxiii 



Robert C . Euler 1 


Durward L. Allen 5 

James E . Deacon and Maxine S . Deacon 9 


Charles Douglas 21 

Richard G. Wiegert 27 


Edward DeBellevue, Howard T. Odum, Joan Browder and George Gardner 31 


John E. Randall, Helen A. Randall and Alan H. Robinson 4 5 


Richard A. Dirks 49 


Wilfred D. Logan and F.A. Calabrese 57 

J. Robert Stottlemyer 65 


Warren F. Steenbergh and Charles H. Lowe 71 


Ronald I. Miller and Larry D. Harris 79 


A.R. Weisbrod 83 

L.G. Abele and Edward F. Connor 89 


Larry E . Morse and Jane I . Lawyer 95 

L.K. Thomas, Jr 97 

Jill Baron, Christine Dombrowski and Susan Power Bratton 101 



D.J. Frederick, L. Rakestraw, C.R. Eder, R.A. Van Dyke, B.J. Griewe 

and M.A. Anderson 107 

L. Rakestraw, D.J. Frederick, C.R. Eder, R.A. Van Dyke, B.J. Griewe Ill 


Carol A. Jefferson 115 

Barbara Ann Coffin 119 

Donald J. McGraw 133 

Richard H . Hevly 151 

Phillip L. DeBord 159 


Ronald H. Hofstetter and Frances Parsons 165 

Frances Parsons 171 


Richard Stalter 177 

Edward E. Dale, Jr. and James W. Gibbons 183 


S.V. Krupa, R.J. Kohut and J. A. Laurence 189 


Walter L. Loope and Neil E. West 195 

Ingrid Olmsted 200 


William H. Thomas 200 


W.H. Moir, F.D. Hobson, M. Hemstrom and J.F. Franklin 201 


R.W. Fonda 209 

Donald B . Lawrence 213 


John Ewel and Louis Conde 225 

Douglas D. Piirto 235 

C.T. David, D.A. Tilles and D.L. Wood 239 


David R. Stevens 241 



Loren D. Potter 247 




Steven W. Carothers, Stewart W. Aitchison and R. Roy Johnson 253 

David J. Schmidly, Robert B. Ditton, William J. Boeer and 
Alan R. Graefe 261 


Leo F. Marnell 269 


Kenneth C . Chilman and John Burde 275 


Aelita J. Pinter 279 

R.W. Frenzel, E.E. Starkey and H.C. Black 287 

Tim W. Clark and Tom M. Campbell 293 


John E. Comely, Hugh H. Genoways and Robert J. Baker 297 

Kenneth L. Diem and Garth S. Kennington 301 


A.R. Weisbrod, W.F. Stevens and G.E. Nordquist 307 

A.R. Weisbrod and J. A. Dragavon 315 

Walter M. Tzilkowski and Frederick F. Knowlton 319 


Richard C . Chapman 323 


Rolf O. Peterson 329 


Audrey J. Magoun 335 

Robert W. Aho and Peter A. Jordan 341 


Thomas C. Collins 349 


Philip C. Shelton 353 

Mark L. Shaffer 357 


James K. Baker 365 

Ronald R. Keiper 369 


Charles L. Douglas and Christopher Norment 373 


Milford R. Fletcher 385 

W. Calvin Welbourn 387 


Clifford S. Crawford 393 

L.M. Ehrhart 397 


David J. Morafka 401 

L.N. Carbyn 409 

James D. Lazell, Jr 415 


David R. Stevens 421 


Charles W. Fowler and William J. Barmore 427 

Clay Dean 435 

David M. Leslie, Jr. and Charles L. Douglas 439 

Henry E. McCutchen 443 

Edwa rd C . Murphy 449 

Douglas H . Chadwick 451 

D.M. Swift, J.E. Ellis and N.T. Hobbs 457 


John R. Haldeman 461 


Roland H. Wauer 469 


Robert L. Neill and Therese M. Allen 475 

Therese M. Allen and Robert L. Neill 479 


James A. Kushlan 483 

Kenneth L. Diem 489 

E.E. Starkey and R.A. Schnoes 497 

Frank B. Isaacs and Norman F. Sloan 501 

Dale L. Taylor 509 

Robert C . Eckhardt 513 

William E. Southern, Francesca J. Cuthbert and Stephen R. Patton 519 


Douglas W. Larson 525 

Richard L. Meyer and Laura L. Rippey 531 

M.S. Ogra, A.C. Sims and G.S. Rai 539 

Jessie S. Ortiz, Stephen A. Souza and William L. Levine 545 


Dana L. Abell 553 

R.C. Mathews, J.D. Sinks and E.L. Morgan 559 

Dennis M. Kubly and Gerald A. Cole 565 


Carl Widmer 573 

Michael Soukup 577 


Luther A. Knight, Jr. and Jack Herring 585 

F.O. Hoffman and J.R. Donaldson 591 

Dana L. Abell 595 


Royal D. Suttkus and Glenn H. Clemmer 599 

K.H. Seethaler, C.W. McAda and R.S. Wydoski 605 

Gerald R. Smith, Robert Rush Miller and W. Daniel Sable 613 

Boyd Kynard and Robert Garrett 62 5 

Clark Hubbs and John G. Williams 631 

Robert Wasem 637 

William F. Trumpf, Eric L. Morgan and Ray Herrmann 643 

Michael H. Hof f and Eric L. Morgan 649 


B.F. McPherson, G.Y. Hendrix, Howard Klein and H.M. Tyus 653 

Gary E. Davis 657 

TO JULY 1974 
Thomas W. Schmidt 665 

Daniel K. Odell 673 


Elizabeth H. Gladfelter and Rosemary K. Monahan 679 




R.G. Layton, W.C. Malm, K.D. O'Dell andW.R. Willis 683 


Marvin H. Wilkening 687 


Gary M. Ahlstrand and Patricia L. Fry 691 


James K . Agee 695 

Keith A. Yarborough, Milford R. Fletcher, Joseph McGown and 
Patty Fry 703 

R. Herrmann, E.L. Morgan and R.L. Green 715 


Robert Giegengack, Elizabeth K. Ralph and Alan M. Gaines 719 

Brian White 727 

Judith A. Schiebout 737 

William J. Fritz and Lanny H. Fisk 743 


William F. Rittschof, Albert B. Dickas and John J. Fisher 751 

Maurice L. Schwartz 757 


Karl F. Nordstrom, Norbert P. Psuty, James R. Allen, Lynn R. Sherman 
and Jonathan R. Pawlow 761 


Stephen P. Leatherman 769 

Richard W. Travis and Paul J. Godfrey 777 


Ervin G. Otvos, Jr 781 

Christopher C . Mathewson and Robert M. McHam 787 


John A. Barnett and Raymond Herrmann 791 


B. Hallet 795 


William 0. Field 803 


Garry D. McKenzie 809 

Richard P. Goldthwait and Garry D. McKenzie 815 


J.C. Harksen, D.K. Olson and E.A. Gordon 823 


Robert C . Lindquist 827 


Wayne L. Hamilton 835 

Alan D. Howard and Robert Dolan 845 



Yvonne G. Stewart 853 



Pamela C . Magers 859 

Don P. Morris 865 


James A. McDonald 869 

T.W. Mathews and E.H. Neller 873 

Roger E. Kelly 875 


Natalie B. Pattison 879 

AND #2 (13Bil9) 
James W. Stoutamire 881 


Gary A. Wright 887 



Adrienne B. Anderson and C. Vance Haynes, Jr 893 

W . James Judge 901 

David A. Breternitz 907 

Lawrence L . Loendorf 911 


Robert C . Euler 917 


Elizabeth Andrews and Kathryn Koutsky 921 


Harvey M. Shields 925 

Merry Allyn Tuten 927 


Dennis B. Fenn and Raymond E. Burge 929 


Christy G. Turner II, Richard L. Bradshaw and William J. Burke 935 


John W. Weymouth 941 

Toni Carrell 949 


John E. Ehrenhard 955 


J.E. Ingmanson and M.A. Colyer 959 

John L. Cotter 961 



Kent Downing and Terry Jo Thompson 96 5 

Ken L. Casavant and James C. Barron 971 

Richard M. Schreyer and Joseph W. Roggenbuck 975 

Kenneth C. Chilman and Marvin Brown 985 

Robert H . Becker 991 


Glenn E. Haas 999 


F.P. Noe and Kirk Elifson 1003 

Michael D. Grimes, Thomas K. Pinhey and Daniel E. Campos 1013 


Elwood L. Shafer 1021 


D.Q. Brodie and Susan F. Taylor 1023 



Gregory Schalliol, Margaret Morgan, John Reynolds, Robert Schiller 
and Thomas Fake 102 7 


Jan W. van Wagtendonk 1033 


Jan W. van Wagtendonk 1039 

Michael M. McCarthy, Anne F. Frondorf and Lewis S. Albert 1041 

Harvey Fleet, Kenneth Muth and Nancy Fries 1049 


R.H. Giles, Jr., A.R. Tipton, T.L. Sharik, G.J. Buhyoff and 

K.A. Argow 1051 


Robert H. Giles, Jr., A.B. Jones III, A.R. Tipton and T.L. Sharik 1061 

R. Gerald Wright 1067 

R. Gerald Wright 1077 


James L. Sherald 1081 

Joan G. Ehrenfeld 1085 

James L. Sherald 1089 

G.A. Rai, M.S. Ogra, L.Y. Yatsu and J.D. Tallant 1091 


J.R. Parmeter, Jr., M. Srago, N.J. MacGregor and F.W. Cobb, Jr 1097 

J.C. Patterson and J.R. Short 1101 


J.R. Short and J.C. Patterson 1107 


Jay S. Angle, Duane C. Wolf and John R. Hall III 1111 


James C . Patterson and John R . Short 1115 

J.R. Short and J.C. Patterson 1116 

Stephen P. Leatherman 1119 

Charles E. Olmsted 1125 

D.A. Shinn 1135 

Ingrid C. Olmsted 1143 


Dieter Mueller-Dombois 1149 


G . Bradford Shea 1155 


Gary S. Waggoner 1161 


Albert E. Radford and J. Dan Pittillo 1165 


George A. Pet rides 1173 


Clifford W. Smith 1179 

R.T. Brown, R.M. Linn, W.L. Kowalski, W.C. Larsen and R.A. Schulz 1183 


William H. Moir and William M. Lukens 1189 


Milton C. Kolipinski 1201 




W.D. Cocking, E.E. Baxter and S.L. Lilly 1205 



Robert E. Martin and Arlen H. Johnson 1209 


Bruce W. Jeske and Collin D. Bevins 1219 


Collin D. Bevins 1225 


James K. Agee and Harold H. Biswell 1231 

R.W. Fonda 1239 


Robert A. Janke 1243 


William L. Bancroft 1253 

Matthew G. Hickler and Susan Power Bratton 1261 

Daniel 0. Holmes 1267 

Bruce B. Moorhead and Edward S. Schreiner 1273 

Ernest Hartley 1279 

John Lemons 1287 

William L. Kowalski 1293 

Steven H . DeBenedetti and David J . Parsons 1305 


David J. Parsons and Steven H. DeBenedetti 1313 


Richard V Giamberdine, Alex R. Carter, Douglas D. Faris, 
John T. Austin and Joel V. Kussman 1319 

Robert M. Linn 

It is my great pleasure to welcome you to this First 
Conference on Scientific Research in the National Parks, 
dedicated to George M. Wright who was the first Chief of the 
Wildlife Division of the National Park Service from 1930 to 
1936, and whose personal dedication, enthusiasm and knowledge 
were largely responsible for the initiation of scientific re- 
search and the practice of ecological principles in the man- 
agement of the national parks of the United States. 

I welcome not only the conferees, but also the Natural 
Science Advisory Committee for Fish and Wildlife and Parks, 
and the Southwest Regional Advisory Committee, both of whom 
have meetings adjunct to this Conference and to whom I extend 
a most cordial welcome to attend any and all Conference ses- 
sions as their schedules allow. 

National Park Service, Michigan Technological University, 
Houghton 49931. 



Dr. Richard Trumbull, Director 
American Institute of Biological Sciences 

It is my pleasure to welcome you to this first conference on 
scientific research in the national parks. I find this most ap- 
propriate for a setting in New Orleans, a pleasant reminder that 
so much of the national park scene of interest to us falls with- 
in the Louisiana Purchase. Further, one is reminded of the 
journal of the LeMornes on their way from Montreal to settle in 
New Orleans. This journal was filled with notes on the abundance 
and beauty of both flora and fauna. They commented on six-foot 
catfish in the Susquehanna and paid little attention to trees 
under four feet in diameter. 

Today's awareness of our responsibility for conservation of 
our heritage must be heightened by a greater understanding of it. 
Science, then, becomes the handmaiden of stewardship. There are 
many benefits to be derived from this gathering, in fine papers, 
meeting of old friends and discovery of mutual interests. We 
have hopes that even as your disciplinary interests are awakened, 
you will also gain a better perspective of the interdisciplinary 

We would have you experience that which the early painters of 
the emerging American scene a century ago tried to express. It 
is said that Thomas Moran (1837-1926) painted the geysers and 
canyons so beautifully that one of his pictures was purchased by 
Congress and is given credit for the later founding of the nation- 
al park system itself. 

"These painters were all of the nature-loving and heartier 
school. They saw big, they painted big, they appealed to those 
who thought big. If man is meant to have dominion over the earth-- 
this earth, this America — then man must be splendid indeed .... 

"But then, American artists have always felt that if we study 
our wondrous landscape, immerse ourselves in it, we can create our- 
selves anew. We can be powerful. Or we can be serene. Perhaps 
we can even be both."* 

"Eden and Awe: The American Landscape," by Mary Ritchie, 
Delta Air Lines Magazine, Sky, November 1976, pp. 20-23. 


Finally, I am pleased to indicate that others appreciate 
your participation in this conference. Quoting a letter from 
the White House: 

"Of particular interest to the White House are the innova- 
tive steps being taken by the American Institute of Biological 
Sciences and the National Park Service to mobilize the scienti- 
fic community for some effective, integrated use of the vast 
scientific and educational resource that is the National Park 

"The White House extends congratulations to the American 
Institute of Biological Sciences and to the National Park Service 
for their joint efforts to promote parks as a national research 
resource. " 

Let us do our best to accomplish that in the programs to 
follow. We are happy to have you all here. 


Assistant Secretary of the Interior for Fish and 
Wildlife, National Parks, and Outdoor Recreation 


Theodore W. Sudia, Chief Scientist 
National Park Service 

If my introductions this morning tend to seem somewhat 
pedestrian, it's because there appears to exist some confu- 
sion in the minds of those of us who are not Park Service as 
to who is what to whom in our little hierarchy. I am deter- 
mined to clarify the basics. 

Nathaniel P. Reed is the Assistant Secretary of the Interior 
who is charged with responsibility for Fish, Wildlife, and 
National Parks. His science overview thus is broader than the 
Park System, including all Wildlife Refuges as well as the ex- 
tensive scientific research that goes on within that other 

The Secretary of the Interior looks to Nat Reed for advice 
on how developments elsewhere in the Department will affect the 
natural land and water systems within the fish and wildlife and 
park systems. This is not an easy job in a Department that has 
been aptly nicknamed "the dustbin of government." The conflicts 
are many and horrendous, and Secretary Reed has had some tough 
on-the-job training in the past 7 years. 

But he did not come unprepared. He had served two governors 
of two different political persuasions as Florida's consultant 
on matters of environmental concern. Environmental conflicts of 
interest do NOT tend to shrink when viewed from the state and 
local level. 

He is here today to wish us well as we launch this historic 
"first" in science deliberations. We are most pleased to be 
welcomed by Assistant Secretary of the Interior for Fish, Wild- 
life, and National Parks, Nathaniel P. Reed. 


Secretary Reed's remarks are not available for the Proceedings 
publication. However, at their conclusion, he awarded the Depart- 
ent of the Interior's Distinguished Service medal to William B. 
Robertson, Jr. , and read the following citation, signed by Interior 
Secretary Thomas S. Kleppe. 

"For distinguished service, to William B. Robertson, Jr. , in 
recognition of his outstanding contributions to the National Park 
Service in the field of ecology and natural sciences. 

"Dr. Robertson's 21 years of Government service have resulted 
in an enhanced understanding of the complexities of the ecosystem 
of Everglades National Park and the whole of South Florida. As a 
research biologist he pioneered in the use of fire to restore 
natural ecosystems and contributed to the understanding of the role 
of fire in natural resource management. Dr. Robertson in his long 
and distinguished career has become a world expert in the ecology 
of the sooty tern, the wood ibis, and the colonial nesting birds of 
South Florida. He has served as advisor to numerous natural re- 
source managers in South Florida and was frequently called upon to 
advise the Office of the Secretary and the State of Florida on mat- 
ters relating to the ecological well-being of Florida. Dr. Robert- 
son's long list of publications which have appeared in national 
and international journals is ample testimony of a long and distin- 
guished scientific career. In 1971 Dr. Robertson had conferred 
upon him the rank of senior scientist. He was elected to the 
Society of the Sigma Xi, the Phi Sigma Society, and was made a fel- 
low of the American Association for the Advancement of Science. Dr. 
Robertson's standing in the national and international scientific 
communities reflects the professionalism with which he has long 
served the Government and the reputation by which he has earned 
outside support for Service policies. In recognition of his out- 
standing contributions as a scientist and researcher, William B. 
Robertson, Jr., is granted the highest honor of the Department of 
the Interior, the Distinguished Service Award." 



Gary Everhardt 
National Park Service 

I want to begin this conference by thanking you for extending 
me the privilege of welcoming you. I represent management of the 
National Parks and my presence on this platform speaks more elo- 
quently than any words. It says "We've come a long way." 

Ten years ago I would not have been here. And if I had been, 
I wouldn 1 t have looked out into the audience and seen the number 
of managers I see here today. Ten years ago you would have been 
scientists talking to scientists. And if that fact implies any 
fault, that fault lies as much in my court as it does in yours. 

But 10 years ago is only a point in time, and today we are here, 
together. And so I submit that we can start this conference on 
a note of hope, for we have already made progress. 

I do not necessarily assume that science and management are 
superior to science itself. But as long as parks must be man- 
aged, then I propose that science and management is better for the 
parks than either is, by itself. 

Just as I don' t assign all blame to science for the separate- 
ness of the past, neither do I award all credit to scientists for 
the progress we demonstrate here today. 

There have been important changes in attitudes. 

Managers today — and those who are moving into positions to be- 
come managers tomorrow — are a different breed. They are more 
sophisticated in their overall education and thinking. The new 
manager has seen the results of decisions which would have been 
different had more facts been available at the time the decision 
had to be made. 

Today's manager doesn't want to be remembered tomorrow for 
some monumental blooper. (Or even for some natural area blooper.) 

I realize that science must be free, but I'm going to risk see- 
ing you flinch by saying that perhaps we haven't managed science 
enough. And before your howls rise to the heavens, let me add that 
we managers haven't had enough voice in the choice of research and 
control of the information flow that results. 


I do not mean to say that managers are competent to design 
specific research projects. That is your field and we leave 
it to you. But managers are much more aware than scientists of 
the park problems that in themselves are demanding management 
answers. We have a duty to make you aware of these problems — 
where and when and how they may rear their heads. 

And then you, in your infinite scientific wisdom, can go out 
and slay our dragons — or at least pin them to the mat so that we 
can slay them. 

Management and science have in the past been two separate 
rivers, each contained within its own comfortable banks, both 
flowing through the National Parks but seldom mixing. Manage- 
ment knew science was there, and presumably was "up to something." 
But when the tough decisions had to be made, often times the 
appropriate results of scientific research simply were not avail- 
able. Maybe science had the answers, but management never asked 
the questions. Often because it didn't know whom to ask or what 
to ask. 

Today I'm in a choice position to ask questions. I look at 
this address as an opportunity to throw out some questions—ques- 
tions that constitute a challenge to both science and management. 
I hope that in the course of your deliberations here this week you 
can come up with some answers. Managers have shown a willingness 
to reassess science in the parks. I would hope that scientists 
are willing to look at what they do within the context of manage- 
ment problems. 

Essentially management's role is to evaluate information and 
make decisions. We need the information that only science can af- 
ford. And public awareness of environmental imperatives is forcing 
management's hand. If we dig a sewer to serve our visitor loads 
and happen to hit a wrong soil or rock — if we project an airplane 
runway or punch an access road through a sage grouse strutting 
ground, we aren't just criticized any more — we're hauled into the 
courts — those hallowed halls of justice where good intentions get 
no brownie points and ignorance is no excuse. 

The legal and political ramifications of managing the National 
parks are helping management hear science better. 

We need better integration of scientific research into the mis- 
sion of park management, and I am here today to put the stamp of 
management approval on that objective. 

We are dealing here with two distinct areas of responsibility 


which must somehow, and soon, be fashioned into one strong 
stand. Management has certain duties; science has others. 

It is the duty of management to perceive and assess cor- 
rectly the problems faced in the parks. Having identified 
these problems, it is management's continued responsibility to 
pose appropriate questions for research. 

It is the duty of research to move quickly at this point — 
to acquire the facts, to answer management's questions, to 
identify the various possible alternatives, to inform about 
the consequences of each, and them to assess scientifically 
the impact of implementing the action plan chosen by manage- 

We need your information all along the line. When we are 
developing our planning requirements, we need to know what our 
basic natural resource consists of — how its processes operate — 
and what visitors will expect of it. The clear focus of poten- 
tial and of expectations by the public should be made at this 
point, and it is a joint job for both science and management. 
It carries an important interpretive element too. Vie must know 
what a proposed area can deliver, and the public must be in- 
formed as to what they can expect to gain from its operation as 
a park site. 

The next step — a resource management plan — should stem from 
a realistic assessment of the natural potential and the human ex- 
pectations. Information is needed at this point to fine-tune one 
or the other and bring the two into agreement. Your natural and 
social science research can help us decide which is the more 
easily manipulated and how to achieve harmony. 

Visitor use plans — another element of the general management 
process — also depend on what we know about an area and the people 
it will serve. As managers we have a duty to be specific in the 
questions we ask, and to ask them within a time frame that allows 
you to do your best in furnishing answers. 

When all these decisions have been made and implemented, we have 
a park area and we have people using it. We come now to visitor 
interpretation. This is where science first entered the National 
Parks. It was naturalists, interpreting the uniqueness of our great 
natural areas, that were science's first representatives within the 
National Park Service. The role of science in interpretation is at 
least as important today as it was in the beginning. And now it 
has an additional task — the job of interpreting the impact on park 
areas and visitors of science's own stepchild, technology. The 
National Park System was founded on the bedrock of nature and history, 
and if we don't have our interpretive stories straight, we aren't 
doing our job. 


Finally, we come to the DCPs — the development concept plans. 
Although our area roots are all embedded in the past, we are being 
dragged into the future willy nilly — and again we are faced with 
decisions. How do we deal with continuing park processes and 
changing visitor demands? 

Science has a continuing task to keep our information current 
in both areas of science—natural and social. As managers we must 
remain constantly aware that parks and visitors are never set in 
any one condition for all time. Everything about a park area is 
in a constant state of change. Parks are process, and so are vis- 
itors, and so most of all is the interaction between them. We need 
all the information we can get on both parks and visitors, and we 
need to be alert to their ever changing natures. 

One of the roles we share is the effort to anticipate problems. 
This forward-looking role is recognition that there will be a 
future, and that parks are not just for us, now, but also for future 
generations. A large part of meaningful science in parks is an at- 
tempt to project parks into the future. 

Today park managers are still heroes. Help us continue to be the 
guys in the white hats. I think we're moving in the right direc- 
tion, but let's quit doing it inches at a time. Let's make it meters- 
kilometers at a time: Park reorganization has put us into the mode 
of acceleration. Let's take a close look at how we can accelerate 
the integration of management, science and interpretation. Train- 
ing, career opportunities, lateral movement between the fields — all 
of these possibilities are open and being scrutinized right now. 
Help us with your input. 

Still unresolved is the issue of how much science should be done 
in-house and how much by contracting with outside research capabil- 
ity. A judicious mix of both seems best suited to our present Ser- 
vice needs. The NPS science staff should be developing the basic 
long-range understanding of natural area processes. 

It should provide a clear, stored and retrievable picture of 
these processes. Then, within this larger context, specific manage- 
ment needs can be defined and urgent short-term research projects 
answering the needs can be let out on contract to universities or 
private research groups. 

It is clearly necessary that we move toward enlarging our science 
staffs in the more seriously threatened parks. The kind of ongoing 
scientific monitoring that an in-house staff can provide is impera- 
tive if we are to cut our casualties. Redwoods, Indian Dunes, Ever- 
glades — our critical list is getting uncomfortably long. 


The direction we are taking is to move massively, one step at 
a time. Everglades is our first giant step. I recognize that 
science in the parks as a whole is understaffed and underfunded. 
But to bring it up to a realistic standard across the board would 
be financially prohibitive and perhaps premature. A tested over- 
all game plan is needed, and what we are doing in Everglades is 
being done with an eye to eventual Service-wide application. 

At Everglades, we have committed ourselves to the establish- 
ment of a model science and resource management program. A build- 
ing will be staffed by both resource managers and research scien- 
tists and dedicated to making these two functions strongly singular. 

We hope and expect that this approach will bridge some of the 
existing gaps between resource management and research. We have 
created new science positions in the park and a science advisory 
board outside it. The integrity of the in-park program will thus 
be subject to the critical review of outside eyes. We have com- 
mitted 450,000 additional dollars to the 1977 phase of this thrust, 
and added $2 million to the fiscal year 1978 budget. 

I look upon the Everglades project as the first of what must be- 
come a System-wide assessment of our major natural areas. As we 
strengthen and tighten our research and management programs we ex- 
pect to become increasingly successful. What we learn in the Ever- 
glades will help us formulate plans for the next major area to go 
the line for similar treatment. 

I have recently set up a Task Force, headed by Deputy Director 
Briggle and including the NPS Chief Scientist and two members of 
Secretary Reed's Science Advisory Committee — Drs. Durward Allen and 
A. Starker Leopold. I have asked them to look at the overall NPS 
science effort — where it came from, where it is today, and where it 
needs to go in wedding science to resource management. The Task 
Force has an April 1 report deadline. 

Two new opportunities have arisen recently for science to contri- 
bute to our Service mission straight across the board — in planning, 
design, management, operations, maintenance and interpretation. One 
of these is the Bicentennial Land Heritage proposal and the other is 
the energy crisis. Our efforts in response to both must clearly be 
tied to our science capability. 

As we add to our System under the Land Heritage Act, and as we 
manage and interpret that system in energy-conserving ways, we have 
a chance to pioneer new territory-- to discover better "fits" be- 
tween man and the rest of nature. We can show the way ourselves, 
by scientifically wise management and maintenance practices, and we 


can educate others by sound interpretation. 

Advances in energy conservation already are underway. The new 
Lovell Visitor Center at Big Horn Canyon is solar-heated; at Gate- 
way a windmill will soon be powering electric vehicles and the vis- 
itor center; and at Cuyahoga NRA in Ohio we're designing a park that 
will be entirely energy self-sufficient. Our energy conservation 
efforts are being coordinated by the Chief Scientist with the Direc- 
tor of Management Services and his Departmental conservation commit- 

In conclusion, my specific questions are these: (1) How can man- 
agement cut down on its mistakes? (2) What can technology do for us 
as park managers? (3) How can information management be used to launch 
needed research in a time frame that will coincide with management's 
need for answers? (4) What science information storage and retrieval 
technology is available and (5) how can it be adapted specifically 
to tie science and management into a closer, more productive working 

These are our challenges and our opportunities. Neither science 
nor management has fully measured up to its responsibilities so far. 
Perhaps too much time has been spent contending about the job and 
too little in attending to it. People roll up their sleeves to do 
one of two things — fight or work. We've had enough of fighting. As 
long as our sleeves are rolled up, I suggest we get to work — mesh our 
efforts — fulfill our responsibilities. 

Again, I appreciate the opportunity to welcome you to this gather- 
ing, and with your help at every field level I can assure you we can 
accomplish our goals for science and research in the National Parks. 
Thank you. 


Robert M. Linn 1 

It's been great meeting all the old faces again; and even 
greater to see all the young scientists who are using the 
national parks as laboratories for their research. The success 
of this Conference is indebted to many, especially the following: 

Lorraine Tucker and Carol Chisholm of the American Institute 
of Biological Sciences, who worked very hard (and succeeded ad- 
mirably) in making a myriad of logistic decisions and arrange- 

All the good folks at the registration desk; 

The Audio-Visual guys from Tulane University--I have person- 
ally never witnessed audio-visual arrangements that have been done 
so well and without mishap as the ones at this Conference; 

The Program Committee members and session chairmen; and 

Those who could not obtain institutional or governmental funds 
to travel here, but came anyway — sometimes at great personal sacri- 

We look for and welcome any suggestions for improving the next 
Conference on Scientific Research in the National Parks, which 
ought to occur in about 3 years. 

National Park Service, Michigan Technological University, 
Houghton 49931. 



Stanley A. Cain 
University of California, Santa Cruz 

When I accepted the invitation to make some summary re- 
marks, I had no idea of the magnitude of the program — three 
and a half days of program with many concurrent sessions, all 
on results of scientific research projects in the National 

It is a far cry from the time 15 years ago when I wrote a 
plea for Research by the National Park Service at the request 
of the Chairman of the Advisory Board, and a far cry from my ser- 
vice as Assistant Secretary of Interior from 1964 to 1968 when I 
continued to stress the importance of scientific research in the 
National Parks and direct participation by scientists in the Service. 

Therefore although it is a shock to heft the bulk of the ab- 
stracts of papers of this Conference, you can imagine how gratifying 
a shock it is. It is good to know how many scientists are doing re- 
search in the Parks, and especially what a relatively high percen- 
tage of them are Park employees or are receiving Park assistance. 

I have done some counting. I will not impose the details on you, 
but merely approximate the statistics. 

I find that there are about 350 separate contributions to the 
program and that, importantly, more than 300 of them are based on re- 
search in the parks. About 200 of the contributions have biological 
subject matter. More than 100 of them are by Park Service employees, 
including supported work at cooperative centers. Looking more closely 
at the involved Park personnel, I find that they come from throughout 
the Service: 54 individuals from a total of 18 different parks, 20 
from Research and Service Centers, 21 from 9 University Cooperative 
Research Centers, and 47 from 12 different administrative centers in- 
cluding Washington, Regional Offices and the like. Park Service em- 
ployed scientists have participated side by side with scientists of 
non-federal institutions. Papers came from the USDA, eight units in- 
cluding six different. National Forests. Contributions have come from 


three Atomic Energy Commission laboratories, five U.S. Geological 
Survey bases, as well as from the Army Corps of Engineers, the 
Bureau of Land Management, the Fish and Wildlife Service, Bureau 
of Outdoor Recreation, and the Public Health Service. In addition 
to the Park Service personnel and those of other federal agencies, 
I find representation on the program by five different state 
agencies, five private museums, five non-governmental organizations, 
six private corporations, one foreign agency, and four private 
citizens. Looking at the program another way, the contributors have 
official connections with 106 institutions of higher learning in 39 
different states. Counting all kinds of contributions, every state 
in the Union is represented in the program. 

In a sense this, the First Conference on Scientific Research in 
the National Parks, is a new phenomenon. After a very long gesta- 
tion period, the concept is born fully mature. How ever logical the 
use of the parks is for appropriate research, it is comparatively 
recently that the National Park Service itself has employed signifi- 
cant numbers of well-qualified investigators not only to help work 
out management problems but to increase knowledge of natural condi- 
tions and phenomena of which many could be studied only in Parks. 

How great has been the change in management's attitude toward re- 
search could be documented by many of us, but such comments are now 
pointless. Today we can and should express our appreciation to the 
Director, to the Chief Scientist, and to the Co-Chairmen for the Con- 
ference, one representing the National Park Service, the other the 
American Institute of Biological Sciences, as well as to the many 
cooperative Park Superintendents. 

Having listened to as many papers as possible, I have heard a 
fair cross-section of the presentations in the several subject- 
matter sections with, however, a bias toward ecological matters. 
This hasn't been much of a bias, however, because ecological rela- 
tionships are central to the studies of most of the participants. 

Many of the papers might have been reserved for presentation be- 
fore other forums such as the American Association for the Advance- 
ment of Science and the American Institute of Biological Sciences 
as well as the meetings of the individual specialized sections. I 
have become convinced that the authors were attracted to this con- 
ference because the Parks contain many of the best natural labora- 


tories in America; but there seems to be a second strong reason, 
a desire to be useful as well as grateful to the Service. To 
this end a surprisingly large number of the studies are directed 
toward helping solve management problems. This is a double thrust. 
Many of the studies provide information that helps administrators 
see their way through complex and sometimes conflicting situations 
toward supportable decisions. Also many of the studies yield infor- 
mation that is suitable for the public that likes to know more 
about the attractions of the parks, hence for enriching museum and 
visitor-center presentations, campfire talks, etc. 

In summary of a summary, I trust that it is not amiss for me 
to draw certain conclusions and make recommendations. You have 
given me the opportunity and I have tried to see the Conference as 
a whole and integral part of the Service's concern. 

1. A large percentage of the papers on the program were matured 
recently and based on data gathered during the last 2 or 3 years. 
This suggests, and I therefore recommend, that the National Park 
Service plan to follow this conference with similar ones at 3- to 
5-year intervals, for there promises to be an abundance of useful 
and interesting information that should be published promptly. 

2. Although many if not most of the present papers would find 
ready publication elsewhere, they should appear as a Proceedings of 
this and subsequent conferences. 

3. If the Service is to get full benefit from research in the 
parks, as many top-level management persons as possible should con- 
tinue to attend future conferences. The "whole show" has a tremen- 
dously greater impact than the papers that deal with a specific 
interest or problem. 

4. As soon as the Proceedings become available and NPS personnel 
have had a chance to peruse them, it would be useful to hold in-park 
seminars on pertinent subjects as a follow-up to this conference. 


Read over New Orleans TV stations on November 8 and 9, 1976 

This week New Orleans is hosting a quiet little conference that 
holds important and hopeful seeds for the future. The National 
Park Service and the American Institute of Biological Sciences are 
meeting with scientists and park managers from all over the United 
States to take their first joint look at scientific research in the 
National Parks. 

To quote NPS Director Gary Everhardt: "This is our real, tan- 
gible way of saying we recognize that the National Park System is 
just a part of the total world biosphere. The problems we face in 
parks are only smaller versions of the problems faced by the entire 
world. " 

The great natural area parks in our System are there because they 
exemplify the highest quality environments that Earth has been able 
to achieve in the course of natural evolution. In today's energy- 
conscious society, we can look to these parks for vital clues as to 
how the great natural power systems of Earth function — the sun and 
wind and tides that keep matter cycling through all the living 
systems that support us. 

National parks are superb examples of energy and matter in dynam- 
ic balance. The Park Service has a double mission--to serve and to 
preserve. In its management role, the National Park Service must 
know as much as possible about how the natural systems work; and it 
must arrange to fit people and their needs into these systems with 
the least possible damage to the environment. 

The research that goes on in parks is providing baseline informa- 
tion about the power and matter flow of the biosphere — that thin, 
fragile security blanket that makes all the difference between Earth 
and the other planets that circle our Sun. 

Fitting people into high quality environmental systems and pre- 


serving the quality — that, after all, is what we are essentially 
seeking to do, worldwide. To provide for people, efficiently and 
economically, taking full advantage of the natural energies that 
everywhere abound, is the challenge of the day. 

As members of a highly technical modern society, we tend to pin 
our faith on technology. But Lewis Mumford told us 10 years ago 
that no technology ever invented by man is anything but a faithful 
imitation or a hideous caricature of some bio-technical process of 
nature. Instead of piling on another layer of technology, perhaps 
we ought to be stepping out of the way. 

If parks can plug into the sun to heat their visitor centers; 
if parks can plug into the wind to power batteries for smooth, 
silent electric buses; if parks can plug into swamps and cypress 
domes to avoid building tertiary treatment plants, then the rest of 
the world should be taking a page from the Park Service book. 

As usual, Shakespeare said it best: "Tongues in trees, books in 
the running brooks, sermons in stones and good in everything." The 
parks are eloquent teachers — especially when the lessons arise from 
our having stubbed our toe on nature. 

New Orleans would do well to pay attention this week as the 
National Park Service holds its first conference on scientific re- 
search in the National Parks. 



Robert C. Euler 

I am most pleased and honored to have this 
opportunity to address a few brief remarks to 
this plenary session of the First Conference on 
Scientific Research in the National Parks . As 
an anthropologist and scientist who has long 
been concerned with the topic of this conference, 
I hope that I can present to you some cogent and 
productive thoughts based upon this experience. 

At the outset, some clarification of my posi- 
tion -- and, hence, my biases — may serve to 
place my remarks in a clearer perspective. My 
tenure in the National Park Service, aside from 
1-1/2 years stint as a Ranger in 1948, is only 
entering its 3rd year. I am new to the internal 
operations of the Service, and there is much I 
do not know. Being relatively new, however, 
perhaps I can offer a relatively new viewpoint. 

Secondly, I consider myself as an anthro- 
pologist to be scientifically oriented, not only 
to the social sciences, but especially to the 
natural sciences. In this, I should like to pre- 
sent a functional, holistic approach to science 
in the national parks. Natural and social or 
(cultural) resources must be viewed as functional 
interdependents , not separately. To put this 
another way, anthropology, at least cultural 
anthropology, in my view, is a part of natural 
science in which we study and attempt to under- 
stand human beings and their society and behavior. 
Since human societies no less than other animal 
societies are an integral part of nature, we can 
study them scientifically at all times and places, 
We do this functionally and holistically , relat- 
ing each aspect of behavior to the other, to the 
whole, and to the natural environment. Human 
behavior thus viewed is a system, or perhaps in 
some instances a series of subsystems interact- 
ing with the natural environment. Scientific 
resources in our national parks, be they bio- 
logical, geological, or cultural must be viewed 
systemically , not as disjointed parts. 

Thirdly, I am a firm believer in basic re- 
search. Many national parks and monuments, pro- 
tected as they are, unique as they are, present 
ideal research potential. Basic research by 
both Park Service and other professional scien- 
tists should be given every possible support and 
encouragement. This means that research projects 
should not necessarily be tied to immediate 
management needs but should be supported for 
their own sake. As has been voiced elsewhere, 
the significance of scientific resources in our 
parks cannot be evaluated in a research vacuum. 

Intimately related to research is the neces- 
sity for an adequate and rapid publication sys- 
tem. I have forgotten now who it was who phrased 
that old saw: "Not publishing research is like 
winking at a girl in the dark; no one knows 
you ' re doing it. " 

Important, cogent research in the parks, or 
for that matter anywhere else, can be carried 

National Park Service, Grand Canyon National 
Park, Arizona, and Southern Illinois University, 

out only by scientists who possess what some have 
termed "the scientific spirit." It was Francis 
Peabody who remarked: 

A greater gain to the world than all 
the growth of scientific knowledge is the 
growth of the scientific spirit with its 
courage and serenity, its disciplined con- 
science, its habitual response to any dis- 
closure of the truth. 

The attainment of this scientific spirit is 
not easy. But, basic ingredients to its achieve- 
ment include intellectual honesty, academic 
stimulation through colleagues and libraries, 
freedom of thought, inquiry, expression, and 
flexibility of activity. As G. Spencer Brown, in 
his book, THE LAWS OF FORM (1969), said a few 
years ago: 

To arrive at the simplest truth . . . 
requires years of contemplation. Not 
activity. Not reasoning. Not calculation. 
Not busy behavior of any kind. Not read- 
ing. Not talking. Not making an effort. 
Not thinking. Simply bearing in mind what 
it is one needs to know. And yet, those 
with the courage to tread this path to real 
discovery are not only offered practically 
no guidance on how to do so, they are ac- 
tively discouraged and have set about it in 
secret, pretending meanwhile to be dili- 
gently engaged in the frantic diversions 
and to conform with the deadening personal 
opinions that are being continually thrust 
upon them. 

This may seem to some to be a most unusual and 
stringent view, but I quote it because the lives 
of too many of us, scientists and others alike, 
are far too structured and mechanical. 

In sum, my philosophy stresses the interre- 
lationships of the sciences, the undeniable r 
strength of well planned basic research, the 
dissemination of data through timely publication, 
and the acquisition and maintenance of the 
scientific spirit. 

How are these thoughts reflected in National 
Park Service divisions? In looking at and pos- 
ing questions primarily about my own discipline 
and the roles of scientists within the National 
Park System, I perceive several areas where in- 
novations may be warranted. By way of intro- 
duction to this reflection, let me tell you of 
en interesting experience I had not long ago. 

I had the privilege of speaking to a group of 
Park Service superintendents attending a seminar 
in resource management in one of our regions. 
Some of the stated objectives of that seminar, 
it seems to me, are germane to our meeting here. 
For example, one objective was "to reestablish 
our understanding of the basic values of the 
National Park System." To my way of thinking, 
cultural and natural — that is, scientific -- 
resources are some of the most important basic 
values of the system -- the cultural ones being 
irreplaceable, non-renewable. 

Secondly, the formulators of that seminar said 
that they wanted "to restate the priority position 
which management of resources holds among total 
service responsibility." Really, no one should 
argue with a position that suggests that if these 
resources are not to be irretrievably lost, their 
management deserves a very high priority indeed. 

Thirdly, the participants of the seminar were 
asked "to reevaluate the specific responsibilities 
of park managers in relation to the management of 
these basic values." I submit that realistic 
understanding of the nature of scientific resources 
and planning for needed research is such an impor- 
tant responsibility. 

Finally, the conferees were asked "to provide 
an atmosphere which will allow the reorder of area 
priorities." I remarked then, and I reiterate it 
now, that attitudes in our minds come about 
through interest, through understanding, through 
excitement, and through effective communication. 

Let us look at f unctionalism in science for 
a moment. In my own field of anthropology — 
especially in its archaeological aspects -- re- 
search is impossible without the functioning co- 
operation of botanists, zoologists, surficial 
geologists, hydrologists , and a host of other 
types of scientists. Human societies interact 
with the plants, animals, soils, water — indeed 
with all of nature (unless I totally misunderstand 
ecology) , and the archaeologist cannot begin to 
comprehend human behavior without close inter- 
digiting with these other disciplines. The effect 
of human activity on land and water must be of 
some concern to them also. Hence, I am disturbed 
-- no, puzzled would be a more apt word — when I 
see what I perceive to be a general dichotomy 
between the archaeologists in the National Park 
Service and those who are more frequently termed 
"scientists" or "researchers." I almost missed 
an opportunity to attend this conference because 
announcement of it was routed to "the researchers" 
and not to me because a secretary on our staff did 
not think a conference on science would be of con- 
cern to an anthropologist. In fact, I am told that 
anthropology was at first not considered an inte- 
gral part of this conference for somewhat the same 
reasons . 

Western Regional Archaeologist Roger Kelly, in 
his paper, "Strange Bedfellows?" prepared for pre- 
sentation at this conference, has made some tel- 
ling points in focusing on what should be mutual 
concerns of scientists from different disciplines. 
He mentions, as just one example, the problem of 
the effect of controlled forest fires on surface 
archaeological materials in the area to be burned. 
At Grand Canyon, anyway, biologists carried out 
"controlled" burns without any consideration for 
the impact they might have upon archaeological 
resources. Without any prior research data upon 
which to base an opinion, one biologist informed 
me that the fire would not have any adverse ef- 
fect. I have just trudged through ankle deep 
ashes and charred stumps to locate previously jn- 
recorded cultural resources to make at least an ex- 
postfacto assessment of his statement. 

It should be a concern of managers to urge the 
dimming of the distinction between natural and 
cultural areas as much as is practicable in our 
parks. The administrative separation of scientists 
such as biologists and anthropologists should be 
eliminated. And scientists, if they are dynamic, 
should be in the forefront in proposing such a 
unification . 

I have already alluded to my contention that 

research seems to be far down the priority list 
in our parks and that research immediately un- 
related to administrative programs appears to be 
apathetically received if not in considerable 
disfavor. There seems to be ample funding for 
research when the administrator perceives a pro- 
blem, little if any when the scientist perceives 
one. Let me again give one example. The Con- 
gressional passage of Public Law 93-620, the 
Grand Canyon National Park Enlargement Act, of 
January 3, 1975, transferred 185,000 acres of 
national park and national forest land to the 
Havasupai Indian Reservation. I do not begrudge 
this; certainly the Havasupai have needed a 
larger land base. But, long before this transfer, 
federal managers of the lands in question had 
been charged, by Executive Order 11593, with the 
responsibility of conducting a cultural resource 
inventory thereon. This had not been done. It 
is safe to say that more than 90 percent of the 
affected acreage had never been visited by an 
archaeologist. Furthermore, Executive Order 
11593 clearly states that federal lands will not 
be "transferred" prior to the completion of a 
cultural resource inventory. But, the lands 
were transferred in what appears to be a clear 
violation of the Executive Order. What may be 
worse is that no one seems disturbed -- not the 
Park Service, not the Forest Service, not the 
Bureau of Indian Affairs, AND NOT THE ARCHAEO- 

About a year ago, Kenneth Boulding of the 
Institute of Behavioral Science at the University 
of Colorado, published in SCIENCE (Boulding 1975) 
an editorial entitled "Truth or Power?" It 
dealt with the relationship between science and 
politics. If I may take the liberty of substi- 
tuting the word "administrative" for "political" 
-- in some quarters, there is little if any 
difference -- Boulding 's editorial would read in 
part as follows: 

The relationship between the scientific 
and the [administrative] communities is one 
of constant mutual frustration. There is a 
feeling on both sides that each ought to be 
able to help the other. The [administra- 
tive] community is constantly faced with 
making what it thinks are at least impor- 
tant decisions. Every decision involves 
the selection among an agenda of alternative 
images of the future, a selection that is 
guided by some system of values. The values 
are traditionally supposed to be the cher- 
ished preserve of the [administrative] 
decision-maker, but the agenda which in- 
volves fact or at least a projection into 
the future of what are presumably factual 
systems, should be very much the domain of 
science. Bad agendas make it much harder 
to make good decisions and if the decision- 
maker simply does not know that the results 
of alternative actions will be, it is 
difficult to evaluate unknown results. The 
decision-maker wants to know what are the 
choices from which he must choose. It is 
not surprising, therefore, that there is a 
demand for a one-armed scientist . . . 
without that infuriating other hand. 

I hope, parenthetically, that you look upon me 
as neither one-armed nor infuriating. 

Perhaps through mutual dialogue that aims at 
eliminating value judgments and ethnocentric 
biases on the parts of both the scientist and the 
administrator we may achieve desired and har- 
monious results that will benefit the parks a"-d 
the American people. 

Turning to the third concern I wish to discuss, 
that of the dissemination of information, I sense 
around me a pride in the results of publication. 
At least, there is a great deal of talk about it, 
especially on the part of the scientists. I am 
afraid that we are deceiving ourselves. The 
xeroxing of a few copies of a report for the files 
is not publication, and I suspect that hundreds of 
reports of potential scientific import lie buried 
in files where they remain unread by the scientif- 
ic community. The National Park Service should 
be playing a leadership role in this area. Again, 
turning to my own field, I see a minimum effort 
to publish research results; and a frustration 
with the imponderable mechanisms through which it 
may be accomplished. In the western United States, 
two regions of the U.S. Forest Service are publish- 
ing timely and inexpensive offset produced studies 
based upon research in their areas. The Park Ser- 
vice should take immediate steps to establish 
editorial boards to judge the merit of and facil- 
ities to actually publish the results of scientif- 
ic studies in its areas. This could be done not 
only in regional offices, but also in larger parks 
such as Grand Canyon. I realize that some of this 
is being done but it is not being accomplished to 
the degree of volume or timeliness that is necces- 
sary . 

I turn now to the final and perhaps the most 
important of the concerns that I wish to talk about 
-- that of the establishment and maintenance of 
the scientific spirit. Here, too, as with the 
other topics I have mentioned, I direct my at- 
tention to managers and scientists alike. This is 
also a realm that is fraught with frustration in 
both quarters. Administrators have little con- 
ception of how scientists work; scientists feel 
that administrators expect them to perform like 
mechanical clerks . 

Administrators, through their often apathetic 
or even hostile attitudes toward science can do 
much to destroy the morale that comes with the 
scientific spirit. One brief example that quickly 
comes to mind is the recent experience of a 
National Park Service scientist who was invited to 
present a paper about his in-park research at an 
international meeting. Knowing that Service funds 
were not available for support, and deciding to 
use personal funds for travel following the meet- 
ing, the scientist made but one request. That was 
that annual leave not be charged against the time 
he was actually in attendance at the meeting. His 
request was denied in the regional office and he 
was told that he was not to consider himself a 
Park Service representative at the meeting. To 
add to this demoralizing position, the scientist 
was further told that the international gathering, 
incidentally celebrating the centennial of its 
existence, could not be very important or the 
scientist would have asked for the necessary 
travel funds to attend. I think that most would 
agree that that is one good way to dim the spirit 
and reduce the effectiveness of personnel. 

to the scientific spirit, you are in the wrong 
business. If you cannot accept the responsi- 
bility and discipline that comes with the flexi- 
ble freedom to work as one chooses, you are not 
a scientist. 

My primary suggestion here is for the estab- 
lishment of any incentive program for scientists 
who demonstrate that they merit it. I do not 
mean monetary incentives, but rather I strongly 
believe that the Park Service should emulate 
the academic community and provide regularly 
scheduled sabbatical leaves for deserving pro- 
fessional scientists. In this way, further 
training, refresher courses, time for writing, 
and a stimulation of that scientific spirit 
could be obtained. 

Allied to this, I believe that the Park Ser- 
vice should enter into cooperative agreements 
with colleges and universities to provide short 
courses, of less than a semester's duration, for 
Service scientists. This could do much to up- 
grade the capabilities of many of our intel- 
lectual vegetarians. 

Let me indicate to you, finally, that these 
are but my ideas, voiced, I hope, constructively. 
They run the usual risk of over-generalization 
and I am sure that my criticisms do not apply 
to all. Yet, if they stimulate one administra- 
tor or one scientist, my effort will be repaid. 

It is your move administrators! You can make 
a choice and not leave to chance the important 
collaborative decisions that must be made if 
science in the parks and science and the parks are 
to flourish. Look realistically at what scientif- 
ic research can offer you, boost its priorities, 
take time to understand scientific attitudes, and 
positive results will accrue. 

And it is also your move scientists! You, too, 
can make the choice rather than gamble about the 
contributions that you can make. Speak out, take 
an active part with responsible aggressiveness, 
seize the scientific spirit, and be excited about 
your careers . 

Together we can enhance our lives and those of 
untold numbers of visitors to the parks in the 
future, at the same time, we can make lasting 
contributions to knowledge. 


1975. Truth or Power? 
The Laws of Form 



Geo. A-len and Unwin. 
KELLY, ROGER E. 1976. Strange BEdfellows? 

Points of in-park cultural and natural resources 

research. Ms., First Conference on Scientific 

Research in the National Parks. 

London : 

Many scientists also have their professional 
utility reduced by being shunted into administra- 
tive positions that trained and efficient clerks 
could be performing. Yet it must also be admitted 
that some have come into federal service because 
of their inability or unwillingness to compete 
with their peers in a more professional sphere. 
I see many individuals who have been trained in 
the sciences simply vegetating in their secure 
positions. To those people I would say, if you 
are an 8 to 5 person, or one who would prefer to 
sort pencils or construct paper clip fly traps 
rather than to sublimate your professional life 

Durward L. Allen 

A widely recognized and traditional purpose of 
our National Park System is to preserve represen- 
tative samples of North America's choice scenery 
and natural land and water types. It is also an 
acknowledged ideal that the great wilderness parks, 
most of which qualify under the "national park" 
category, should be complete and self -operating 
ecosystems in whatever degree that is possible. 
According to longstanding criteria, a "national 
park" is large in size, has several outstanding 
features, and is inviolate against exploitive 
uses. This distinguishes it from monuments, which 
can have a single major preservation feature and 
be more subject to disturbing influences. Classic 
examples of national parks are Yellowstone, Mount 
McKinley, Great Smokies, and Everglades. Good 
examples of monuments are Saguaro, Organ Pipe, 
and Death Valley. Obviously, there are areas 
worthy of preservation in units of other types, 
for example bogs or tracts of virgin timber in 
recreation areas. 

There are also land reorganizations that could 
upgrade the National Park System. One that comes 
immediately to mind is the possibility of a 
Sonoran Desert National Park that could include 
Organ Pipe and the Cabeza Prieta National Game 
Range on the Mexican border in Arizona. There 
are possibilities for an international park, in 
case the Government of Mexico wished to partici- 
pate by adding their outstanding caldera area in 
the Sierra del Pinacate immediately to the south. 

As short-handed as the Nationa 
has been in the past decade, it is 
if everything is not getting done. 
Alaska's problems are taking up si 
is none, and that will continue fo 
However, we have an historic need 
go away, and I propose to help kee 
as an objective for attention as s 
comes feasible. We have no major 
truly represents the Great Plains, 
grasslands in general have been ne 

1 Park Service 
At present 

ack where there 

r a while. 

that will not 

p it before us 

oon as it be- 

park that 
and the 


This is not to ignore the excellent examples 
of grassland vegetation, with some appropriate 
animal species, in such areas as Wind Cave 
National Park, Theodore Roosevelt National 
Memorial Park, Badlands and Devil's Tower National 
Monuments, the National Bison Range, and Wichita 
Mountains and Valentine National Wildlife Refuges. 
But grassland is not just grass, and its true 
beauty and productivity cannot be appreciated 
unless it is complete with its natural complement 
of forbs and animal life. Grassland animals 
include the large and small, the vegetarians and 
the carnivores, and they need space to utilize 
their range and pursue their seasonal behaviorisms 
in the natural manner. As true grassland parks, 
all of the areas mentioned fall short on size and 
the fauna they contain. Setting aside any area 
in the west for public use always is complicated 
by the indigenous attitude that all grass should 
be grazed by domestic livestock. 

1 — 

Department of Forestry and Environmental 
Sciences, Purdue University, West Lafayette, 
Indiana 47907. 

To avoid misunderstanding, it should be made 
clear that I am not discussing the tall-grass 
prairie that has been under consideration in the 
Flint Hills of Kansas. This is a particular type 
of grassland, essentially a phase of what Weaver 
(1954) and others designated the "true" prairie. 
It lay adjacent to the forested lands to the east, 
and on the lower, more moist sites was character- 
ized by the tall grasses such as biq bluestem 
(Andropogon gerardi), Indian grass (Sorghastrum 
nutans), switchgrass (Panicum virgatum), and 
sloughgrass (Spartina pectinata). Especially on 
uplands, it was dominated by many of the not-so- 
tall midgrasses, most notably little bluestem 
(Andropogon scopar i us ) . 

The true prairie developed some of the most 
productive agricultural soils on earth, and it is 
understandable if such areas are almost totally 
under the plow. In the Flint Hills, rocks at the 
surface restrict cultivation, and the grazed 
prairie vegetation could easily be restored to 
something like its primitive condition. This 
prospective park would not be of a size to accom- 
modate the larger mammals in a wild state and 
qualify as a complete ecosystem. However, it 
could be an attractive sample of the tall-grass 
vegetation, with many of the smaller animals that 
belong there. I assume it would be 50-60 
thousand acres. 

As a generality, we can say that the eastern 
prairies occupied a region of forest climax, with 
a moist subsoil, and they were maintained in a 
grass "disclimax" (disturbance climax) condition 
by recurring fires. Some of them are hardly 
recognizable today because the cessation of fires 
after the beginning of agriculture permitted 
forest invasion of uncultivated lands. However, 
the areas of deep, black, level soils in northern 
Indiana, northern Illinois, and Iowa are obviously 
of prairie origin. The land is mostly in a one- 
year "rotation" to corn, and crop ecologists say 
they are putting back all they take out. In 
another quarter-century, we may be able to judge 
that more accurately. 

For conceptual purposes I find it convenient 
to regard the true climatic grassland, if there 
is such a thing, as beginning (at the mid-states 
latitude) beyond the Mississippi River in eastern 
Nebraska, where the tall grasses largely disappear 
and several kinds of midgrasses become dominant. 
Westward it includes certain short grasses; hence 
it is the "mixed prairie", the most extensive of 
our grassland types. Nearly all of these types 
include a great array of largely perennial and 
deep-rooted forbs that made the open lands an 
ever-changing garden of flowers through the season. 
Few people of today have any idea what the mag- 
nificently attractive prairie flora was like. 

The mixed prairie is underlain by a dry sub- 
soil and a mineral hardpan or lime layer usually 
at a depth of 4 to 5 feet in the east, where annual 
rainfall is about 25 inches. As we proceed west 
into regions of lower rainfall the lime layer 
rises to within a foot or so of the surface. 

I regard the mixed prairie north of Nebraska 
as offering the best possibilities for a Great 
Plains National Park. Historically the region 
was highly productive in both diversity and 
quantity of animal life and offers great interest 
in terms of seasonal change. It probably was the 
prime habitat for both buffalo and the Plains 
Indians of a century and a half ago. 

The principal dominant grasses of this region 
are the needle grasses (Stipa spp.;, western 
wheatgrass (Agropyron smithii ) , Junegrass 
(Koeleria cristata) , the dropseeds (Sporobolus 
spp.;, little bluestem and side-oats grama 
C Bouteloua curtipendula). Short grasses that 
occupy a lower level in the open stands are 
mainly blue grama (B. gracilis ) and buffalo grass 
(Buchloe dactyloides) , the latter being replaced 
northward by several sedges (Car ex spp.;. Pro- 
gressively westward onto the high plains, the 
midgrasses give way to the short grasses. Weaver 
and Albertson (1956) , confirming earlier work by 
Clements (1916), and Shantz (1923), observed that 
short grasses gained ground over the midgrasses 
as a result of heavy grazing. The same factor 
reduced the short grasses in turn and permitted 
such inferior forage as big sage (Artemisia 
tridentata) , rabbitbrush (Chrysothamnus spp.;, 
and prickly pear (Opuntia spp.;, to take over. 
Sagebrush ranges are now particularly prevalent 
in Wyoming and Montana. 

Earlier I indicated some indecision as to 
whether a "true" grassland climax exists. This 
is a question hinging largely on whether one con- 
siders fire as an intrinsic attribute of climate. 
In semi-arid lands that can support a vegetation 
thick enough to burn, fire is a universal occur- 
rence. "Natural" lightning fires are difficult 
to separate from man-caused fires, since men 
have been setting fires for many thousands of 
years. I tend to follow the thinking of Sauer 
(1950) and Stewart (1951), that practically every 
grassland would be invaded by woody vegetation in 
the absence of fire. On our northern prairies 
the evidence is convincing in the stands of 
chokecherry (Prunus virginianum) , wolfberry 
C Symphor icarpus occidental is ) , buffaloberry 
(Shepherdia argentea) , silverberry (Eleagnus 
commutata), sumac (Rhus spp.;, and other woody 
plants that have escaped burning for a number of 
years. Favorably situated ravines, other rough 
topography, and upwind-facing waterfronts are 
taken over by coverts that bespeak the effect of 
fire dynamics on local sites. Wells (1965), 
pointed out the role of topography, soils, and 
bedrock formations in preserving scarp woodlands 
in the grass country. He concluded that there 
are species of woody plants that can survive in 
practically all areas where grass does not produce 
enough thatch to burn effectively. 

Our grasslands develop their most salient 
characteristics and greatest stability on large, 
relatively level open spaces, where they have the 
earmarks of a climax. I am satisfied to regard 
them as such with the understanding that fire and 
the normal complement of animal life were built- 
in maintenance mechanisms. The park we are dis- 
cussing would need to incorporate these features, 
although it would not be objectionable if it 
included some rough lands illustrating the diver- 
sity of the region. For this same reason, it 
should be far enough west to grade into the short 
grass and sagebrush habitats of the high plains. 

The idea of such a park is not new; Shelford 
(vide 1940) and others were discussing the need 
for grassland preservation early in the century, 

and he is reported as suggesting a great plains 
park to the National Park Service as early as 
1930 -- a move that had support in the Ecological 
Society of America. In 1940 Cahalane described a 
plan for a Great Plains National Monument in 
southwestern South Dakota and northwestern 
Nebraska. I was not aware of Cahalane 's detailed 
plans at the time I carried out my own field 
reconnaissance during several summers of the 
early sixties, and it gave me some reassurance 
to discover that I finally centered my own in- 
terest in the area he recommended. It is open 
and undeveloped, although private and public 
lands are all grazed. Both Shelford (1940) and 
Cahalane decided that a million acres would be 
needed to create a truly representative and self- 
contained park, a view to which I subscribe. 

Extensive field surveys have been carried out 
in all parts of the country by personnel of the 
National Park Service in determining the park 
needs of the country. I am told that they too 
are in general agreement that the area where 
South Dakota, Nebraska, and Wyoming come together 
should have first consideration in locating a 
great plains park. It could turn out to be a 
tri-state area. There are butte lands across 
the Wyoming border that could be an attractive 
inclusion, and the possibility of taking in a 
corner of the Nebraska sandhills should not be 
neglected. A major consideration in the plan 
offered here is that the park must avoid any 
towns or primary highways. If the area in ques- 
tion proves to be unfeasible, then the northwest 
region of South Dakota should be investigated. 
It seems to offer outstanding opportunities, 
although I have been through it only once. 

Let us visualize an area roughly 26 by 60 
miles, with the long axis from east to west. It 
would need to be west of the Pine Ridge Indian 
reservation; the exact shape would depend on 
topography and other considerations. The park 
probably should center around a height of land, 
since it would have to be surrounded by a barrier 
fence -- a development that might cost as much 
as a county road. Much of this duplicates 
Cahalane's original plan. 

The fence wo 
and wolves, whi 
unmanaged assoc 
park. The outf 
would require s 
and these might 
conform to the 
interior border 
outside it shou 
access road. 

uld need to be proof against bison 
ch would be, in their wild and 
iation, a unique feature of the 
lowing drainageways , commonly dry, 
ome engineering at the fenceline, 

include waterholes designed to 
scene. The fence would have an 
ing service trail, but on the 
Id be as far as possible from any 

As a preliminary plan, we can 
park would have two entrances — 
connected by a network of strateg 
blacktop, 30 mph, one-way auto tr 
would be many turnouts and rest s 
other developments. Except for a 
and administrative headquarters, 
side public accommodations could 
owned. In concession stands, nat 
illustrating the rich culture of 
Indians should be featured. 

assume that this 
east and west — 
ically located, 
ails. There 
tops, but no 

visitor center 
all of the out- 
be privately 
ive crafts 
the Plains 

This would be a daytime park to which the 
public could be given almost unlimited access in 
their own automobiles. They could witness the 
habits and interrelations of animal life at 
leisure in open vistas permitting every observa- 
tional advantage. Experience indicates that 
nearly all animals would become accustomed to 

slow moving vehicles and largely ignore the human 
presence. Such a park should be closed at night. 
The roads would serve a valuable management func- 
tion as fire breaks for the controlled burning 
that would be required to maintain a healthy grass- 
land community. Under such a regime it is pre- 
dictable that the vegetation would recover its 
primitive aspects and composition and that exotic 
plants now widespread in the region would be 
reduced through competition with adapted native 
species . 

The dominant animals of this grassland, and its 
outstanding attraction, would be the free-living, 
unmanaged plains bison and its only important 
native predator, the wolf. As a rough estimate, 
and depending greatly on location and range condi- 
tions, I would expect that a million-acre fenced 
range might support 8,000 buffalo (about 40 bands) 
and 200 wolves (15 to 20 packs) . Numbers would 
fluctuate with trends in rainfall and other wea- 
ther factors. Pronghorns would supply a portion 
of the food of the wolves, as would elk and deer 
in case wooded ridges and/or stream bottoms were 
included . 

The Great Plains National Park would be a 
primary refuge for scarce and endangered mammals 
and birds, including the black-footed ferret, 
swift fox, prairie falcon, and golden eagle. 
Inevitably, it would support an attractive and 
varied display of grassland birds. Jackrabbits, 
prairie dogs, and many kinds of ground squirrels 
and other rodents would provide food for aerial 
predators as well as such carnivores as the 
coyote, bobcat, and badger. 

and twenties through the urgent requests of the 
late Dr. E. H. McCleery, of Kane, Pennsylvania. 
McCleery bred his wolves and is said to have kept 
the blood lines distinct. 

It needs to be pointed 
be endangered biologically 
on the official list, and 
ly to the plains bison (Bi 
our parks this subspecies 
livestock, subject to annu 
dipping, and brucellosis v 
of nature these management 
by the wolf according to c 
to human judgment. The pi 
one place where the bison 
recover its old, wide-rang 
normal socially organized 
speciation directed by car 
was in primitive times, 

out that a species can 
, even though it is not 
this applies particular- 
son b . bison). In all 
is being managed like 
al handling, culling, 
accination. In a state 

functions are performed 
riteria largely unknown 
ains park would be the 
could live in the wild, 
ing habits, develop 
bands, and have its 
nivore selection as it 

Finding a proper wolf for a grasslands park 
will be a major biological problem. Depending on 
location, the area probably would be at or near 
the presently recognized border-line (region of 
intergradation) between the extinct great plains 
wolf, Canis lupus nubilis, and the northern 
Rocky Mountain wolf, Canis 1. irremotus (vide 
Goldman 1937, Hall and Kelson 1959). Both preyed 
on the buffalo, and the western form irremotus 
evidently still hangs on as a scattered remnant 
in the Wyoming-Montana-Idaho region, where Yellow- 
stone and Glacier National Parks are potential 
sanctuaries. Whether these wolves can be restored 
naturally to a viable population is only a tenuous 
possibility at present, but conceivably public 
attitudes are changing enough to make this 
possible. If it happened, a source of wild stock 
would be available. 

On the other hand, some captive wolves that 
probably are i rremotus or even irremotus x nubilis 
have been held and propagated by Jack and Marjorie 
Lynch at Gardiner, Washington -- the only such 
animals in existence to my knowledge. Jack Lynch 
and others are studying the history of these 
animals, whose progenitors were captured in the 
wild by Biological Survey trappers in the 'teens 

When the current studies 
will be more certain of wha 
inspection of some original 
cates that the animals came 
would make their descendant 
for introduction to a great 
i rremotus are unavailable, 
wolves is also a potential 
precariously situated wild 
stone-Glacier region disapp 

are complete, Lynch 
t he has. Preliminary 
catch records indi- 
from Montana. This 
s priority number one 
plains park, if wild 
This group of captive 
back-up in case the 
wolves of the Yellow- 
ear . 

The wolves at Gardiner, Washington, are in 
excellent condition, kept in large pens and in 
socially compatible groups. Jack Lynch 's study 
of their origin is being aided by the endangered 
species staff of the Fish and Wildlife Service. 
If these animals can be confirmed as irremotus , 
in whole or in part, their preservation is a 
matter of public concern. It should no longer 
depend on the practically unfunded E. H. McCleery 
Lobo Wolf Foundation and the devotion of a few 
people. The Lynch effort is urgently in need of 
assistance . 

Over the earth, for some thousands of years, 
men have been changing grasslands into desert. 
The process continues, although since the drought 
years of the thirties range scientists have 
greatly improved their knowledge of provident man- 
agement in various grazing lands of the western 
states. Commendably, they have also managed to 
get some of our plowed-up semiarid lands back into 
grass before the next drought period . 

However, there is much more progress to be 
made, and one of the primary needs is for control 
areas where mature grassland communities can be 
studied — their natural production, their degree 
and mechanisms of stability. Aside from being a 
superlative area for public observation, a Great 
Plains National Park would have unique scientific 
value. We have already learned that some of our 
research on the behavior and relationships of 
animal life can best be carried out on undisturbed 
and unmanaged lands and waters. This is a major 
value of wilderness, and we cannot say now that 
we have any authentic grassland wilderness. 

Even the park I am proposing would not dupli- 
cate the primitive because of extinctions (plains 
wolf and grizzly, Eskimo curlew) and the presence 
of exotic species. This need not deter an effort 
to achieve what is possible, and although there 
are uncertainties, the result could be very good 
indeed. It must be emphasized that that size is 
part of the ecology of the grassland community. 
A million-acre park (nearly half the size of 
Yellowstone) will be much more than the sum of 
its parts. 

It is a platitude to say that the time to 
establish national parks was long ago -- when 
(bless Yellowstone!) some of it actually was done. 
With some foresight, we could have had a great 
plains park easily in the drought period of the 
thirties, an opportunity that could be repeated. 
The need for recreational open space, scientific 
reserves, and wilderness preservation has been 
increasingly recognized -- as it progressively 
becomes more difficult. With the growth of popu- 
lation, every economic asset on land and water is 
under competitive demand. This casts no asper- 
sions; it simply states that a nation with too 
many people, and more to come, could experience a 
catastrophic mutilation of its natural scene. 

Organizations and individuals concerned for 
the problems and rights of future citizens are 
increasingly embattled to save anything in its 
original form. In many areas where parks are be- 
ing established — especially urban areas — local 
interests must be bought out at prices that are 
possible only because the government can print 
dollars. A park in valuable grazing land undoubt- 
edly would be expensive, but it would not compare 
with some of the other ventures the Congress and 
the National Park Service have undertaken. 

The Great Plains National Park would be like 
nothing else we have, and for once we could start 
in an undeveloped area and apply all that has 
been learned. On at least two occasions recom- 
mendations have been made to the Secretary by the 
National Parks Advisory Board that the National 
Park Service carry out further surveys and devel- 
op plans for the great grassland park that is our 
most obvious "hole" in the nation's program of 
wildland preservation. This much, at least, 
would not be prohibitively expensive, and if it 
were done, I have no doubt that there are many 
members of Congress who would be interested in 
the proposal. 

There is much at stake, and it certainly is 
worth a try. 


CLEMENTS, FREDERIC E. 1916. Plant succession. 
Carnegie Inst. Washington Publ. 242. 

GOLDMAN, EDWARD A. 19 37. The Wolves of North 
America. Jour. Mamm. 18(l):37-45. 

mammals of North America. Ronald Press, N.Y. 
2 vol. 

SAUER, CARL 0. 1950. Grassland climax, fire, 
and man. Jour. Range Mgt. 3(1):16-21. 

SHANTZ, HOMER L. 1923. The natural vegetation 
of the Great Plains. Assoc. Amer. Geog . Ann. 

SHELFORD, VICTOR E. 1940. The need for grass- 
land reservations and grassland research. 
Canadian Field-Nat. 54:5-7. 

STEWART, OMAR C. 1951. Burning and natural vege- 
tation in the United States. Georg. Rev. 41: 

WELLS, PHILIP V. 1965. Scarp woodlands, trans- 
ported grassland soils, and concept of grass- 
land climate in the Great Plains. Science 148 
(3667) :246-249. 

WEAVER, JOHN E. 1954. North American Prairie. 
Johnson Publ. Co., Lincoln, Nebraska 348 pp. 

and ALBERTSON. 1956. Grasslands of 

the Great Plains. Johnson Publ. Co., Lincoln, 
Nebraska 395 pp. 


James E. Deacon and Maxine S. Deacon 1 

Fishes living i 
United States have 
rential currents, 
strates of the Col 
oped humps , appare 
torn; fusiform bodi 
cles and expansive 
swiftly moving cur 
largest minnow in 
pounds; reduced or 
small eyes, charac 
silt-laden streams 
expressed to varyi 
inhabiting Grand C 

n National Parks of the Western 
physically adapted to the tor- 
and shifting, unstable sub- 
orado River. They have devel- 
ntly for stability at the bot- 
es, with slender caudal pedun- 
fins for maximum power in the 
rent; large size, including the 
the world at more than eighty 

embedded scales, and relatively 
teristics of fishes living in 
Most of these features are 
ng degrees by the native fishes 
anyon National Park. 

The most extreme hump is exhibited by the 
humpback chub, Gila cypha; the most extreme fusi- 
form shape by the bonytail chub, Gila elegans; 
and the largest size by the Colorado Squawfish, 
Ptychocheilus lucius. Both above and below Grand 
Canyon National Park the bizarre razorback sucker, 
Xyrauchen texanus , occurs, apparently having 
adapted to less torrential stretches of the river. 
This species occupies some reservoirs of the main- 
stream Colorado. Its decline and extinction in 
many reservoirs built within its native range, 
but its persistence in others, remains a puzzle. 
Some information is available on spawning, feed- 
ing, and relative abundance in some areas. It 
has also been successfully spawned and reared in 
hatcheries . 

At a quite different physical extreme, we also 
have fishes adapted to the extreme ranges of vari- 
ability in salinity, temperature, and quantity 
occurring in the waters of Death Valley National 
Monument. Here are species of the genus Cyprino- 
don living in waters that occasionally form a thin 
film of ice in the winter, and reach maximum tem- 
peratures of about 43°C (109°F) in summer. The 
salinities range from relatively low to concen- 
trations several times that of sea water. Salt- 
encrusting algae form steep-sided, shallow pools 
which eventually build a roof of salt. Occasional 
flash floods may sometimes turn an isolated pool, 
a few meters in length and a few centimeters in 
depth, into a portion of a river over a hundred 
miles in length. Seasonal variations in evapo- 
transpiration isolate some portions of a creek as 
small, warm, increasingly saline pools. In such 
habitats, the importance of surviving a few hours 
during the hottest part of the day becomes criti- 
cal. The zone of resistance (Fry, 1971) or resis- 
tance time (Brett, 1956) therefore, may be one of 
the most important adaptive features of this group. 
On a longer time scale, in the Death Valley region, 
climatic shifts have created at Devils Hole what 
may be the smallest habitat in the world contain- 
ing the entire genome of a vertebrate species. 

Man's competition for water has pervaded every 
aspect of his occupation of the western United 
States. Irrigation practices requirei establish- 
ment of a set of societal rules almost as soon as 
irrigation began. A major portion of our present 
federal and state water law was developed in re- 
sponse to the demands to formalize the right of 

University of Nevada, Las Vegas, Nevada, 89154. 

landowners to the surface and subsurface waters 
on their lands. As the resource became reduced, 
competition for the resource increased and 
society was forced to establish priorities and 
make hard decisions relative to its use. 

During the expansionist period of settling 
the west, society's definition of "good" heavily 
emphasized the values of economic and population 
growth. More recently we have been shifting our 
definition of "good" toward more qualitative 
values. Not surprisingly this shift toward qual- 
ity involves the necessity to mediate through the 
courts, the inevitable conflicts that arise be- 
tween adherents of the expansionist philosophy and 
those who believe that quality of life is soci- 
ety's most pressing need. The reflection of these 
societal values in judgments being handed down by 
the courts is evident in the clear and concise but 
eloquent language of Federal District Judge Roger 
Foley of the District of Nevada. In his Findings 
of Fact and Conclusions of Law which appeared in 
a permanent injunction granted by him on 9 April, 
1974, Judge Foley wrote as follows: "The United 
States has shown that the public interest lies in 
the preservation of this endangered species . Con- 
gress has enacted the Endangered Species Conser- 
vation Act (16 U.S.C. 688aa.), and the Secretary 
of the Interior pursuant to that statute has 
identified the Devils Hole pupfish, Cypr inodon 
diabolis , to be an endangered species. Wit- 
nesses offered by the United States testified to 
the importance of the species to mankind. Con- 
gress, state legislatures, local governments, and 
citizens have all recently voiced their expression 
for the preservation of our environment, and the 
destruction of the Devils Hole pupfish would go 
clearly against that theme of environmental 
responsibility . " 

Defining, interpreting and demonstrating the 
benefits to be gained through practicing environ- 
mental responsibility may be the most important 
contribution the National Park Service can make 
to the future of the U.S.A. The Park Service has 
played a key role in reorienting the administra- 
tion of the waters of the West toward increased 
recognition of qualitative values in the environ- 
ment. The Devils Hole pupfish case stands as a 
landmark! The Park Service will continue to play 
an important role in establishing priorities for 
the use of water in the West . it is currently 
widely assumed that if water is needed for people, 
the people will get it -- somehowl What is less 
well understood is that the people need the water 
for purposes other than drinking, irrigation and 
power. These other qualitative needs are beginn 
ning to be established, and the National Park 
Service is and must continue to play a central 
role in that trend. 

It has been well established that the large 
main-stream dams on the Colorado have caused 
significant detriment to the native fishes 
(Vanicek, 1967; Vanicek and Kramer, 1969; Vanicek 
et al . , 1970; Holden and Stalnaker, 1975). In 
addition to interrupting movements of native 
fishes which may have been necessary for success- 
ful spawning, dams have significantly lowered 
the water temperatures downstream, in many cases 

precluding successful spawning of native fishes. 
In Grand Canyon, the closure and operation of Glen 
Canyon Dam has resulted in a significant lowering 
of water temperatures downstream. Some introduced 
fishes, such as carp, red shiner, and green sun- 
fish have been pushed downstream by the lower 
water temperatures. At present, it appears that 
the introduced forms in Grand Canyon may have been 
more adversely affected than the native forms (Fig- 
ure 1) . Collections of fishes made in 1968 show most 
introduced forms occurring upstream to Lee's Ferry, 
just below Glen Canyon Dam. Summer water temper- 
atures below Glen Canyon Dam dropped from about 
20°C in 1968 to about 10°C following achievement 
of full pool level in Lake Powell, and fish col- 
lections subsequent to 196 8 show some introduced 
forms occurring further downstream or restricted 
to mouths of tributary streams. By contrast, the 
native fishes have not shown such severe down- 
stream restriction although they have de- 
clined somewhat in abundance. This relationship 
should be watched closely in the future since it 
could be quite significant to the survival of the 
threatened and endangered fishes living in Grand 
Canyon. For example, the native fishes may simply 
have longer life spans than the introduced forms 
and, therefore, require a longer period of time to 
show the detrimental effects of lowering water 
temperatures. It is also possible, however, that 
the native species have found suitable spawning 
areas in tributary streams or elsewhere in the 
mainstream and that adults are better able to sur- 
vive in the cold tailwaters than are some of the 
exotic species. Recent collections leave little 
doubt that the lower little Colorado River is im- 
portant to the humpback chub, Gila cypha (Suttkus 
and Clemmer, personal communication; and R.R. 
Miller, personal communication) . 

It may be especially significant that Grand 
Canyon is not being intensively managed to en- 
hance the sport fishery for cold water game fish. 
Thus, the native fishes are apparently competing 
primarily with warm water species that have moved 
upstream from Lake Mead. It is commendable that 
the National Park Service has sponsored studies 
that have developed a reasonably good data base 
on the fishes of Grand Canyon National Park and 
obvious that additional information on life his- 
tory of the native species is needed to develop 
management alternatives that would insure their 
continued survival . 

It seems especially important to examine the 
thermal ecology of the native forms. Where and 
at what temperatures they spawn, pass the fry 
and fingerling stages, and what their thermal 
tolerance limits at all stages of the life his- 
tory are. The intriguing point is that there 
seems to be a possibility that we may be able to 
manage the waters within Grand Canyon to enhance 
the native species. 

Probably the two most striking features of the 
environments in which pupfishes of Death Valley 
have evolved are the highly variable conditions 
of temperature and salinity. Much attention has 
been directed to the constancy of the head spring 
environments. The fact that much tectonic activ- 
ity is, and for at least the past few million 
years has been, occurring in the Death Valley re- 
gion suggests that most head springs are probably 
quite transitory. The differentiation that has 
taken place among the pupfishes, therefore, has 
probably been most influenced by the variable 
conditions of temperature and salinity character- 
istic of the marshes. 

The critical thermal maxima for c. diabolis, 
c. milleri, and various populations of c. 
nevadensis have been determined by Brown and 
Feldmeth (1971), Otto and Gerking (1973), and 
James (1968) to be about 42-44°C. Maxima for 

long term survival probably are near 38°C. Low 
temperature tolerance is less well documented 
but in most cases is near 2°C. Brown and Feldmeth 
(op. cit.) viewed C. diabolis as having evolved 
in a thermally constant environment, similar to 
the present condition of Devils Hole. This inter- 
pretation may be untenable in view of the clear 
evidence of higher water levels in Devils Hole, 
and the evidence developed by Mehringer and 
Warren (1976) of marked changes in water avail- 
ability in the Ash Meadows area over the past 
several thousand years . While the species is 
strongly differentiated and has probably been 
isolated from other pupfish populations in Death 
Valley longer than other forms, its habitat may 
also have been thermally variable. 

Observations of responses of C. n. nevadensis 
to thermal variations at Saratoga Springs, Death 
Valley National Monument (Deacon, 1967, 1968) 
follow closely results of experimental deter- 
mination of CTM. Fish have been seen swimming 
voluntarily at temperatures between 8° and 42°C. 
They occasionally dart quickly in and out of 
areas in the marsh with temperatures up to 44 °C. 
At temperatures below 7°C fish are not seen in 
the marsh but can be collected by seining through 
the soft bottom mud. Measurements of diel acti- 
vity as expressed by catch per trap hour in the 
marsh at Saratoga Springs illustrate the in- 
fluence of temperature on activity. On 25-26 
January, 1967, temperature at the mud-water 
interface in the marsh fluctuated between 15° and 
23°C. These temperatures are within the mid- 
range of thermal tolerance for C. nevadensis. 
Catch rate shows a distinct bimodal pattern with 
peak periods of activity in morning and evening 
(Figure 2) . This is also the basic pattern ex- 
pressed under normal conditions, in the main spring 
pool (Figure 3) where water temperatures fluctuate 
narrowly between 26.5° and 31.0°C throughout the 
year. During summer, water temperatures in the 
marsh commonly exceed 3 8°C during mid-day and mid- 
day depression of activity remains evident, per- 
haps in this instance enforced by high temperature 
(Figure 4) . While no activity occurs at the mini- 
mum winter temperatures, a single daily activity 
peak occurs in the early spring as the water be- 
gins to warm up. This condition is evident from 
data taken on 24 March, 1967 (Figure 5) . Thus, it 
appears that C. n. nevadensis exhibits a bimodal 
pattern of diel activity under moderate temperature 
conditions and that at temperatures below 15° and 
above 38°C, activity is reduced or ceases entirely. 

Similar observations have been made on 
Crenichthys baileyi and crenichthys nevadae (Hubbs 
and Hettler, 1964; Hubbs et al . , 196 7; and Deacon 
and Wilson, 1967) . In those species activity is 
restricted at high temperatures and low oxygen 
concentrations. It seems probable that scope for 
activity will be reduced as either thermal or oxy- 
gen stress is placed on the individual. For ex- 
ample, since reproductive behavior is metabolically 
demanding it will be abandoned as scope for acti- 
vity is reduced. Similarly, since feeding prob- 
ably occurs primarily during peak activity, 
metabolic demands for assimilation will likely 
reach a peak a short time after peak activity has 
occurred. If, as Fry (1971) suggests, the meta- 
bolic cost of assimilation reduces the scope for 
activity, it is reasonable to expect minimum activ- 
ity during a period of food assimilation, 
especially when the CTM is closely approached. 
The same result (reduced activity following feed- 
ing) could also be expected under conditions of 
high salinity if osmoregulatory demands effectively 
reduce scope for activity. These are both poten- 
tially fruitful research opportunities for fishes 
living in osmotic or thermally stressful environ- 
ments, such as occur in Death Valley National 
Monument . 


Salmo gairdneri 

Catostomus latipinnis 

Pantosteus discobolus 

Gila robusta complex 

Ictalurus panctatus 

Cyprinus carpio 

Notropis lutrensis 

Lepomis cyanellus 








River miles 










Salmo gairdneri 

Catostomus latipinnis 

Pantosteus discobolus 

Gila robusta complex 

Ictalurus punctatus 

Cyprinus carpio 

Notropis lutrensis 

Lepomis cyanellus 








Glen Late 

Canyon River mites Mead 


FIGURE 1. Distribution of some native and introduced fishes in the mainstream Colorado 
River in Grand Canyon, 1968 (upper) and 1975 (lower) . 

Defense of territories and presence of breeding 
colors during the summer is common in the marsh at 
Saratoga Springs until mid-day temperatures rise 
above 35°C. No fish were seen defending terri- 
tories or with well-developed breeding colors at 
temperatures above 38°C (Figure 4) . Shrode (1974) 
showed tnat eggs of C. n. nevadensis would develop 
at constant temperatures between 20° and 36°C (but 
not at 38° and 18°C); however, development would 
occur when temperatures fluctuated between 30- 
38°C, 20-28°C and 28-36°C. Clark Hubbs (personal 
communication) has recently shown that Cyprinodon 
eggs will develop up to a constant temperature of 
37°C. Thus, it appears that reproductive activ- 
ity is discontinued at temperatures which rise 
above those compatible with normal egg development. 
Furthermore, the primary spawning period occurs 
during spring and early summer when environmental 
temperatures never exceed limits compatible with 
development. Occasionally during mid-summer, 
environmental temperatures may exceed 3 8°C and 
result in egg mortality in the marsh habitat. 
Shrode ' s data indicate that developmental 
temperature tolerance of C . n. nevadensis is 
unusually wide, a situation that is also true for 
adult pupfish. On the other hand, Gerking (per- 
sonal communication) has recently shown that crit- 
ical thermal limits for ovulation of normal ova 

in this subspecies are quite narrow. Critical 
thermal maxima for young pupfish exceed that for 
adults. Thus, under maximum thermal stress, egg 
mortality will be followed by adult mortality, 
while young fish will survive the longest. Crit- 
ical thermal maxima for two-month-old fish ac- 
climated to 36 °C was determined by Shrode (op. 
cit.) to fall at about 44°C. 

Man induced changes in the Death Valley region 
have been rapid and with far-reaching consequences 
Because all individuals of distinctive species or 
subspecies occupy very restricted habitats in this 
area, the changes have more rapidly affected the 
entire population than has been the case in larger, 
and more diverse habitats of the Colorado River 
system. Illustrative of these changes is the ex- 
tinction in 1958 of both Empetr ichthys latos 
pahrump, formerly occurring only at the Pahrump 
Ranch, Pahrump Valley, Nevada. Irrigation with- 
drawals resulted in drying up the only springs 
containing populations of these subspecies 
(Minckley and Deacon 1968) . Except for two trans- 
planted populations, one at Corn Creek Spring, 
Clark County, Nevada, the other in a plunge pool 
on the Lake Mead National Recreation Area, the 
third subspecies, E. 1. latos, and last remaining 
representative of the genus living at Manse 


(Bowman) Ranch, Pahrump Valley, Nevada, would 
have suffered the same fate during the summer of 
1975 when the last remaining natural habitat 
occupied by the genus Empetrichthys temporarily 
dried up as a result of irrigation withdrawals. 

Except for successful litigation which went 
through the U.S. Supreme Court, a similar fate 
would, in all probability, have been the lot of 
the Devils Hole pupf ish . This species is re- 
stricted to a single limestone cavern in Ash 
Meadows, Nevada. The cavern system was made a 
disjunct portion of Death Valley National 
Monument in 1952 by President Truman. Information 
on population sizes is available since 1967. 
Beginning in April 1972, scuba divers have 
been used to assist in visually counting the 
population throughout the entire depth of 
Devils Hole in which it occurs. The cavern 
system is quite extensive, and is only incom- 
pletely mapped (Figure 6); however, Cyprinodon 
diabolis occurs only in the area of sunlight 

penetration. While i 
26 m the population 
near the surface. Wa 
constant at 33°C, and 
characteristics of th 
constant (Dudly and L 
water surface is abou 
of the mountain in wh 
water surface receive 
time during which sun 
surface changes seaso 
seasonal variation in 

t regularly occurs down to 
is far more concentrated 
ter temperature in nearly 

other physical and chemical 
e water are also nearly 
arson 1974) . Since the 
t 2 3 m below the contour 
ich the hole occurs, the 
s little sunlight. The 
light falls on the water 
nally, resulting in marked 

primary productivity. 

Beginning in 1969, significant variations in 
water level were superimposed on the seasonal 
variations in sunlight as a factor of primary 
importance in controlling population size of c. 
diabolis . These changes have resulted from 
ground water pumping in the nearby aquifer 
(Dudley and Larson, 1974). The purpose of pump- 
ing is for irrigation of pasture and hay crops 
to be used for raising cattle. 


Catch per 
Trap Hour 

Temp.(°C) 45 







Temp, at 
mua surface 

Fish per 










Figure 2. Diel variation in catch per trap hour in the marsh at 
Saratoga Springs, D.V.N. M. 25*26 Jan. 1967 
Water temperature -15 -23 °C. 12 

Catch per 
Trap Hour 






Figure 3. Diel variation in catch per trap hour in the constant temperature 

main spring pool at Saratoga Springs, Death Valley National Monument. 
Water Temp. 26.5-29°C. 

15 Catch per Trap Hour 




No Reproductive 

Critical Thermal 

Temp. 46 


Figure 4 Diel variation in catch per trap hour, reproductive activity and temp, in the marsh at 
Saratoga Springs, Death Vallev National Monument, 18 August, 1966. 



Catch per 
Trap Hour 




Fish per 
Trap Hour 



Temp.(°C) 45 




Temp, at 
mud level 









Figure 5 Diel variation in catch per trap hour in the marsh at 
Saratoga Springs D.V.N. M. 24 March 1967 


Figure 6 Vertical section of Devils Hole, Death Valley National Monument. 







8 0.91 



















1967 1968 1969 1970 1971 

Figure 7. Monthly low water levels at Devils Hole, 1967-73. 

1972 1973 

Water level 




















Figure8. Fluctuations in estimated population sizes of juvenile and adult 
Devils Hole pupfish, April 1972- Dec 1976. 

1 r . 

480 Population 


Y=185+.38X R 2 =57 



Light Intensity 
and Duration 






Figure 9 Light intensity and duration correlated with population size two months later. 

480 Population 


Y=236+30X R'=75 



Index (g 0;/day) 












Figure 10 Primary productivity index correlated with population size one month later. 


On 20 September, 1972, water level in Devils 
Hole reached its historic minimum of 1.19 meters 
below a USGS reference point installed on the 
rock wall above the water surface. This was the 
culmination of a downward trend that became evi- 
dent in 1969 (Figure 7) . Frequent census data 
beginning in April 1972 (Figure 8) shows that, 
beginning from what was probably an all-time his- 
torial low population (of juveniles and adults), 
the population increased to July 1972. Between 
July and September, as the water level fell to 
its lowest point, the fish population also crashed. 
During this period, subsequent and prior data in- 
dicate that a continued population increase would 
normally be expected. The crash, therefore, is 
probably attributable to a reduced carrying capa- 
city resulting from low water levels. At the low 
point, all of the loose gravel and rubble that 
could reasonably be removed, was removed from the 
natural shelf in an effort to increase the avail- 
able habitat. In addition, artifical lights were 
installed in hopes of increasing the primary pro- 
duction. The fish population responded by in- 
creasing its numbers to about 250 by January, 
1973. At that point the artificial lights were 
turned off, and the water level continued to rise. 
Within a month the population had declined. By 
March, probably because of the onset of spawning 
(Minckley and Deacon, 1973), a slight increase was 
noted. Between March and June when the next cen- 
sus was taken, a flash flood carried large amounts 
of debris down onto the natural shelf, and NPS 
personnel, in order to enlarge the habitat, again 
removed the debris. The result was that during 
this period, rather than exhibiting its normal 
spring increase, the population declined. On June 
5, 1973, the court appointed a Special Water Mas- 
ter to oversee a permanent injunction granted on 
that date. The injunction was for the purpose of 
maintaining sufficient water in Devils Hole to 
tend to insure the survival of c. diabolis . From 
that time until July 1976, the water level was 
managed so as not to fall below 1.01 meters below 
the USGS reference point. The result has been 
that since June 1973 there has been a relatively 
regular fluctuation in population size. On 2 July, 
1976, the water level began to be manaqed so as to 
remain above .91m. Actually water levels rose to 
about .82m below the reference point. The popu- 
lation responded by increasing to about 400 indi- 
viduals on September 24, 1976 (Figure 8). 

We believe to tend to insure the survival of 
the species, the minimum population size should 
not fall below 200 individuals. It appears that 
the maximum population is about twice the mini- 
mum. Therefore, a desirable maximum should fall 
at about 400 individuals. 

We have some information (less precise for 
1967-68 than for 1972-76) on maximum population 
sizes prior to pumping and, therefore, can arrive 
at some estimates of the requirements for achiev- 
ing a summer maximum of 400 individuals on the 
basis of regression analysis. Using data avail- 
able on maximum population size and water level 
for 1967, 68, 72, 73, 74, 75 and 76 we find a cor- 
relation coefficient of .86 with an R 2 of .73 at 
the 95% level of confidence. This suggests that 
a maximum population of about 400 individuals 
could be expected at a water level of about 0.73 
meters . 

Data on population sizes for 1972-1976 are more 
accurate as mentioned previously. If we eliminate 
the less accurate 1967 and 1968 data and examine 
only the better data, we again find a high cor- 
relation coefficient of .81 with an R 2 of .65 at 
the 95% level of confidence. Note that the slope 
for this line is somewhat steeper than was pre- 

viously true, indicating a more pronounced effect 
of water level on population size. Again, how- 
ever, we find that a maximum population size of 
about 400 individuals can be expected at a water 
level of about .73 meters below the reference 
point . 

The influence of seasonal variations in light 
duration and intensity on population size of c. 
diabolis is evident from data acquired in 1975- 
76. Pyrheliometer recordings show that, with a 
lag time of about 2 months, population size in- 
creases as sunlight duration and intensity in- 
creases. The relationship is strong with a cor- 
relation coefficient of .76 and an R 2 value of 
.57 at the 95% level of confidence (Figure 9) . 

A third and perhaps even mo 
fluential parameter affecting 
the total primary productivity 
cumscribed by the natural shel 
oped a PPR index based on diel 
content of the water. This in 
ever, is limited to the months 
October since sunlight intensi 
the water surface during other 
cient to stimulate sufficient 
ity to be measureable using ou 
The data do show, however, tha 
of about 1 month, population s 
the primary productivity index 
10) . The correlation coeffici 
R 2 of .75 at the 95% level of 
suggests a relatively high cor 
natural system. 

re directly in- 
population size is 

of the area cir- 
f . We have devel- 

fluctuations in O2 
formation, how- 

of May through 
ty and duration on 

times is insuffi- 
primary productiv- 
r present methods, 
t, with a lag time 
ize increases as 

increases (Figure 
ent of .86 with an 
confidence again 
relation for a 

The primary productivity index is based on 
fluctuations in concentration of dissolved 
oxygen over the natural shelf. As the volume of 
water over the natural shelf increases, a larger 
biomass of plants can be accommodated and, to 
that extent, the volume of water over the natural 
shelf may also influence the primary productivity 
index. In effect, this means that as water volume 
increases, potential for primary production also 
increases. This potential is realized only when 
the sun's rays are able to fall on the water sur- 
face. Total production of the pool thus can be 
maximized only by increasing the water volume 
during a period of increasing light intensity. 
Since typically during the period 1970-76 water 
volumes have been decreasing during periods of 
increasing light intensity, it is evident that 
the decreasing water levels have caused a de- 
pression in the population size that would have 
been reached under more stable conditions . The 
important point demonstrated is that increasing 
sunlight appears to stimulate increased produc- 
tion which in turn results in increased numbers 
of fish. 

The regression formula of primary production 
on fish population size suggests that a produc- 
tivity index of 5.47 grams 62 per day is neces- 
sary to achieve a maximum summer population of 
about 400 individuals. Frequent measurements 
of the primary productivity index from 26 July 
- 3 September 1976 show that total production 
estimates over the shelf ranged between 1.8 and 
5.5 g02/day. The highest sustained productivity 
index per unit volume (1.9 g02/m3) during the 
period occurred on 26-28 August, 1976. If that 
index is held constant (probably possible only 
during periods of maximum sunlight) and water 
level increased, the regression formula shows 
that a productivity index of 5.5 g02/day could 
be expected at a water level of about .82 meters 
below the USGS reference point. 


100 90 


Figure 11 Percentage changes in surface area and volume of water over the natural shelf 
at Devils Hole as a function of water level . 

An examination of the changes in surface area 
of Devils Hole as a function of water level 
(Figure 11) shows that if 0.4 3 m. below the copper 
washer is taken as 100%, .73 represents about 90% 
coverage, .82 represents about 85% coverage, .91 
represents about 70% coverage, and 1.01 represents 
about 40% coverage. In estimated volumes these 
four levels represent 51%, 38%, 25%, and 16%, re- 
spectively. Since a fish lives in three dimen- 
sional space, the volumes represent more realis- 
tically the space available for the bulk of the 
important feeding and breeding activities of this 
species . Our data thus suggest that this species 
would probably fluctuate between about 200-400 
individuals if water levels were maintained at 
about .82 m. , thus providing about 38% of the 
volume and 8 5% of the surface area of the original, 
prepumping pool level. 

As we alter our societal values, competition 
for use of water intensifies because one segment 
of society insists on retaining rights which 
another segment insists must shift to permit ex- 
pression of other values. The importance of ex- 
pressing societal values through legislation, res- 
olution, publication, legal action, establishment 
of National Parks, and any other reasonable and 
legal manner, has been dramatically portrayed by 
Judge Foley's reference to such evidence as being 
of importance in reaching his landmark decision. 
Research in the National Parks is essential not 
alone for the intrinsic value of knowing, but also 
because that knowledge will frequently be essen- 
tial to help guide decisions that give expression 
to society's values. 


BRETT, J. R. 1956. Some principles in the thermal 
requirements of fishes. Quart. Rev. Biol. 31: 

BROWN, J. H. and C. R. FELDMUTH. 1971. Evolution 
in constant and fluctuating environments: ther- 
mal tolerances of desert pupfish (Cyprinodon) . 

DEACON, J. E., ed. 1967. The ecology of Sarasota 
Springs, Death Valley National Monument. Re- 
port to National Park Service. mimeo. 104 pp. 

DEACON, J. E., ed. 1968. Ecological studies of 
aquatic habitats in Death Valley National Monu- 
ment, with special reference to Saratoga 
Springs. Report to National Park Service, 
mimeo. 8 2 pp. 

DEACON, J. E. and B. L. WILSON. 1967. Daily 

activity cycles of Crenichthys baileyi, a fish 
endemic to Nevada. Southwest. Nat. 12(1): 31- 

DUDLEY, W. E. and J. D. LARSON. 1974. Effect of 
irrigation pumping on desert pupfish habitats 
in Ash Meadows, Nye Co., Nevada. U.S.G.S. 
open file report 74-188. Denver, Colo. 142pp. 

FRY, F. E. J. 1971. The effect of environmental 
factors on the physiology of fish. Chap. 1, 
pp. 1-99, In: Hoar, W. S. and D. J. Randall. 
Fish Physiol. Vol. 6. Academic Press, N.Y. 

HOLDEN, P. B. and C. B. STALNAKER. 1975. Distri- 
bution and abundance of mainstream fishes of 
the middle and upper Colorado River basins, 
1967-73. Trans. Am. Fish. Soc . 104 ( 2) : 217-231. 

HUBBS, O., R. C. BAIRD and J. W. GERALD. 1967. 
Effects of dissolved oxygen concentration and 
light on activity cycles of fishes inhabiting 
warm springs. Amer. Midi. Nat. 77:104-115. 


HUBBS, C. and W. F. HETTLER. 1964. Observations OTTO, R. 
on the toleration of high temperatures and low erance 

dissolved oxygen in natural waters by Crenich- nodon) 

thys baileyi. S. W. Nat. 8:245-248. SHRODE, J 

JAMES, C. J. 1969. Aspects of the ecology of on dev 

the Devils Hole pupfish, Cyprinodon diabolis Thesis 

Wales. M.S. Thesis. Univ. Nev. Las Vegas, 61 pp. 

Las Vegas, Nev. 70 pp. VANICEK, 

MEHRINGER, P. J., JR. and C. N. WARREN. Marsh, native 

dune and archeological chronology, Ash Meadow, Dam, 1 

Amargosa Desert, Nevada. pp 120-151. in: Univ., 

Elson, R. (ed.). Holocene environmental change VANICEK, 
in the great basin. Nev. Arch. Sur. Research histor 

Paper No. 6, July 1976, Reno, Nevada. cheilu 

MINCKLEY, C. O. and J. E. DEACON. 1973. Obser- robust 

vations on the reproductive cycle of Cyprinodon al Mon 

diabolis. Copeia. 1973 ( 3) : 610-613 . 98(2) 

MINCKLEY, W. L. and J. E. DEACON. 1968. South- VANICEK, 
western fishes and the enigma of "Endangered 1970. 

Species." Science. 159:1424-1432. Utah a 


G. and S. D. GERKING. 1973. Heat tol- 
of a Death Valley pupfish (Genus Cypri- 

Physiol. Zool. 46(l):43-49. 

1974. Genetic and temperature effects 
elopment of the Amargosa pupfish. Ph.D. 

Arizona State Univ., Tempe, Ariz. 

C. D. 1967. Ecological studies of 

Green River fishes below Flaming Gorge 
964-1966. Ph.D. Thesis. Utah State 

Logan, Utah. 124 pp. 
C. D. and R. H. KRAMER. 1969. Life 
y of the Colorado squawfish, Ptycho- 
s lucius, and the Colorado chub, Gila 
a, in the Green River in Dinosaur Nation- 
ument, 1964-1966. Trans. Am. Fish. Soc . 

C. D., R. H. KRAMER and D. R. FRANKLIN. 
Distribution of Green River fishes in 
nd Colorado following closure of Flaming 
Dam. Southwest Nat. 14 (3) : 297-315. 



Charles Douglas 


There is more information available than we 
know what to do with. Information scientists 
have seen the problem coming for many years, but 
they have not been able to forestall it. The 
problem is here and can only get worse. Before 
we look at the future, let's see what has been 
done and what is being done now. 

The amount of literature related to managing 
and long-range planning for natural resources is 
no exception to the dilemma posed by the enormous 
amount of available information from all sources. 
Coping with the environment is an involved enter- 
prise. Knowledge of the environment is initiated 
from many sources: biology, chemistry, the at- 
mospheric sciences, health sciences, air pollution, 
water resources, civil, electrical, and mechanical 
engineering, energy, legislation, transportation, 
social, management, political, and economics, to 
name but a few. 

Many indexing and abstracting services already 
exist in several areas that would be of interest 
to environmental scientists: Chemical Abstracts, 
Biological Abstracts, Nuclear Science Abstracts, 
Physics Abstracts, Water Pollution Abstracts, 
Engineering Index, Environmental Abstracts, the 
NASA Star, and the Government Research Abstracts. 
There are also other publications relating to 
topics that are of interest to environmentalists 
such as journal articles, congressional hearings, 
documents, theses translations, conference pro- 
ceedings, and monographs. 

With all of this information being generated, 
it is no wonder that the computer has been called 
upon to assist in its storage, manipulation, and 
retrieval. At the present time there are collec- 
tions of machine-readable data produced by both 
the federal government and commercial organizations 
which can be searched for some aspects of the 
environmental literature. 

For several years now, the concept of a total 
environmental information system has been postu- 
lated. In Stockholm in 1972 the United Nations 
Conference on the Human Environment recommended 
an information referral service. Participants in 
this conference suggested "a modest and practical 
tool to tell what information services exist, 
where they are, and how to gain access to them." 
In 1972, William Ruckelshaus, then head of the 
Environmental Protection Agency, commented on our 
uncoordinated or nonexistent information retrieval 
systems and stated that environmental information 
is generated by at least 75 different sources 
within the federal government. 

In 1973, the United States Congress passed a 
bill which allowed the U.S. to participate in the 
United National Environmental Program (UNEP) . 
This program was devoted to the development and 
support of an international system for the exchange 
of information. In 1972, Congress proposed the 
establishment of a national environmental data 
system. This system was to serve as the central 

point "for the selection, storage, analysis, 
retrieval, and dissemination of environmental data 
made available to it by federal agencies, state 
and local governments, individuals and private 
institutions." This National Environmental Data 
System and Environmental Center Act was then 
vetoed by former President Nixon. 

Congress, in 
with an addition 
established a Na 
The Data System 
work of new and 
or computer faci 
various areas of 
a system of inte 
with a central f 
general manageme 
the ancillary so 
usually required 
operation." Thi 

1973, again introduced the bill 
al statement: "There is hereby 
tional Environmental Data System, 
shall include an appropriate net- 
existing information processing 
lities both private and public in 

the United States, which, through 
rconnections , are in communication 
acility for input, access, and 
nt. It shall also include all of 
ftware and support services 

for effective information system 
s bill was also vetoed. 

It has b 
in federal 
mental info 
Dr. John W 
when speaki 
as 1972 tha 
centers to 
in the form 

een said many 
policy making 
rmation system 
Townsend, Jr. 
ic and Atmosph 
ng of a nation 
t this system 
center but an 
provide users 
al data they n 
s and formats 

times by persons involved 
that a national environ- 
should be established. 
of the National 
eric Administration, 
al system said as early 
should not be one big 
interrelated system of 
"with the types of 
eed , when they need it, 
they require." 

The University of Georgia, Athens. 

In the meantime, many federal agencies have 
developed data systems which deal with environ- 
mental information. These systems are to be found 
in the Departments of Agriculture, Commerce, 
Defense, Health, Education, and Welfare, the 
Environmental Protection Agency, and the National 
Aeronautics and Space Administration. 

Some of these systems are extremely active, 
providing information on a regular basis to many 
users: EPA has APTIC (Air Pollution Information 
Center) , NERC (National Environmental Research 
Center) , NEDS (National Emissions Data System) , 
SAROAD (Storage and Retrieval of Aerometric Data) , 
STORET (National Water Quality Information 
System) , SWIRS (Solid Waste Information Retrieval 
System) , ENVIRON (Environmental Retrieval On-Line) , 
and NOISE (Noise Information System) . The 
Environmental Information Systems Office (EISO) 
at the Energy Research and Development Admini- 
stration's National Laboratory in Oak Ridge 
provides information appearing in the Water Re- 
sources, Energy, Toxic Materials, Heated Effluent, 
Mercury, and Nuclear Science data bases. NOAA 
has an Environmental Data Service, called ENDEX 
(Environmental Data Index) . This service pro- 
vides "a rapid referral service to existing 
national and global environmental science data 
files. " 

There are many activities being carried out by 
commercial firms in the area of environmental 
sciences. None of these activities, however, are 
as extensive as the existing federal systems. 

It is now time to look at how the National 


Park Service fits into the information handling 
picture. There is no doubt that the approximately 
300 national park sites have been looked at over a 
long period of time; consequently, large files of 
information have been acquired. This information 
exists in individual parks, in many forms. One 
such form is internal files, usually with a classi- 
fication schedule oriented toward administrative 
functions, rather than project or topic, with the 
result that reports related to a particular subject 
such as beach erosion are distributed throughout 
the files. Since the classification schedule is 
designed for servicewide use, hence is based on 
broad generic categories, large clusters of docu- 
ments in a relatively few categories tend to occur 
at the local park level, making the potentially 
most important reports least retrievable. Infor- 
mation exists also in NPS Libraries in many of 
the parks, some with one classification scheme, 
some with another, and some with no classification 
system at all. Another very valuable form of in- 
formation is in research collections belonging to 
NPS employees. These reports, data, and knowledge 
contained in heads that have been collected over 
the years and moved with the person. Other sources 
of information are inventories of resources done 
by state and local units of government, by major 
conservation groups, and by regional planning 
bodies. There are also extensive data in the files 
of NASA, EPA, USGS, and U.S. Bureau of Sport 
Fisheries and Wildlife. Much valuable information 
is produced by universities and should be incor- 
porated into any information system developed by 
the Park Service. 

There appears to be no overall policy within 
the NPS concerning the generation, distribution, 
organization, storage, and retrieval of all of the 
available information materials. Maps, for 
example, are scattered; some are catalogued but 
most are not; no central inventory exists, so it 
is usually not possible to determine if a particu- 
lar map has been produced or if a later version 
is available; nor is it possible in most cases to 
determine the storage location of a map known to 
exist. Virtually the same situation exists for 
all other types of informational resources gener- 
ated by or for the NPS. 

With Congress declaring the 1970s as the envi- 
ronmental decade, resulting in increased interest 
in the environment, and with the ever increasing 
costs of managing natural resources, it became 
apparent that there was a need for a systematic 
procedure for collecting information to manage 
these resources. It was also apparent that it 
would be necessary to record this collected in- 
formation in an organized manner so that it could 
be used not only for immediate management problems 
but also for long-range planning for the national 
park system. It was recognized that the histori- 
cal record would also be important for determining 
trends and their impact on the parks, the manage- 
ment goals which had been established, and the 
long-term effects on the ecosystem which the parks 
represent. Being cognizant of these things, tne 
NPS commissioned a preliminary planning study for 
the design of an ecological and environmental 
management information system, a system that would 
be a part of its management and long-range planning 
for the natural resources under its jurisdiction. 

The Ecological and Environmental Management 
Information System (EEMIS) has been conceived as 
a very broad and complex project, ranging from 
the handling of quantitative data, such as the 
number of various species of wildlife and their 
migration patterns, to sophisticated modelling 
and simulation techniques for projecting the 
effects of actions before they are taken. 

The planning study suggests that EEMIS include 
information in four broad categories, the same 
four that NPS uses in some of its present inven- 
tories: (1) Physical Characteristics, (2) Bio- 
logical Characteristics, (3) Cultural Character- 
istics, and (4) Socioeconomic Characteristics. 
The first, Physical Characteristics, would in- 
clude such factors as the location and size of 
the area, topography, geology and soils, hydrology 
and climate. The Biological Characteristics would 
comprise principally the plants and animals of 
the area and their interactions with the environ- 
ment. Cultural Characteristics would include past 
and present land use, existing facilities, and 
historical attributes. The Socioeconomic Charac- 
teristics of the region surrounding the immediate 
area should include such factors as visitor ori- 
gins, rural-urban population distribution, eco- 
nomic and industrial bases, age, and racial 
composition which influence to a certain degree 
the kinds of recreational activities and types 
of facilities available. The relative importance 
of these characteristics may vary somewhat from 
region to region, but all need to be considered 
in the collection and evaluation of data for 
effective management. Together these categories 
represent an unparalleled effort to collect and 
relate information about the full range of vari- 
ables which constitute the ecosystem. The large 
size of EEMIS, which will eventually handle in- 
formation for the approximately 300 national 
park sites, makes a computer-based system a ne- 
cessity in order to store, manipulate, and 
retrieve the large collection of data which will 

The EEMIS planning study has now been com- 
pleted, but much time, effort, and of course 
money will be required to make it operational. 
The system design will have to be compatible with 
the operational environment to be accepted by 
and useful to the NPS staff. That is, it must 
support decision-making at the level where it 
occurs--at the individual park for the large parks 
or at the regional offices for smaller parks 
managed from those locations. This may seem 
obvious, but it is exactly the reverse of the de- 
sign concept used for most corporate management 
information systems. In most corporate systems, 
the design facilitates decision-making from the 
top down. If it is to be successful, the system 
implemented for the NPS must be compatible with 
and support the existing environment which is 
basically decision-making from the bottom up. 

In arriving at a decision to recommend the 
building of a global system design to meet the 
NPS information needs, careful consideration was 
given to the enormous diversity of the parks and 
to their requirements or needs which were iden- 
tified. Consideration was also given to the 
operational environment within which it must 
function, the existing information resources, 
the state-of-the-art in information processing 
systems of all kinds, and NPS expertise in and 
experience with formal and informal information 

It has been recommended that the information 
needs of the NPS would best be served through 
an integrated collection of subsystems, rather 
than one monolithic centralized system. There 
are a number of reasons for recommending this 
design concept. First, the wide diversity in 
the NPS data (graphics, bibliographic, narrative, 
quantitative) requires a correspondingly wide 
range of information processing techniques (e.g., 
plotting or drawing, test processing, and sta- 
tistical analysis, to mention only the more 
obvious) . Also, there are a number of 

2 2 

significantly different functions to be served by 
the total system—management summary reporting, 
inventory control, maintenance scheduling, demand 
retrieval, publishing, archival recording, trend 
analysis, and many more. Given the current state- 
of-the-art in file structures and file processing 
techniques, it would be prohibitively expensive, 
if not technically impossible, to accomplish all 
these functions through one monolithic processing 
system operating from one logically organized 
data base. Even if it were technologically pos- 
sible, such a system would probably be so complex 
that it would prove operationally cumbersome and 
conceptually difficult to understand and use. 

The extremely large size of the data collections 
which will eventually result from an NPS-wide im- 
plementation would probably prove unmanageable in 
terms of administrative span of control, especially 
as there is little prior experience within the 
Park Service upon which to build. For this and 
other reasons, phased implementation of identi- 
fiable subsystems over a period of time would 
provide greater probability of success, allowing 
time for education of the NPS personnel both as 
operators and as users of the various components. 
The subsystem approach will also result in shorter 
development periods than would a single, mono- 
lithic system, at least so far as visible progress 
from the users' viewpoint is concerned. It also 
permits prior investments in existing computer- 
based processing systems (e.g., the graphic dis- 
play modules developed for the Great Smoky 
Mountains RBI) to be incorporated either on a 
temporary or a permanent basis. 

Integration of the recommended 
be achieved through a "star" conf 
central subsystem serving a maste 
function. All information betwee 
passes to and from this central s 
is no direct interchange between 
subsystems. A direct interchange 
central subsystems would result i 
control and would make the entire 
technically difficult to operate, 
to be successful, the central sub 
advised of changes in the status 
systems; the likelihood of the ce 
being notified of changes in the 
creases substantially if other pa 
subsystem are allowed to link dir 

subsystems would 
iguration with a 
r coordinating 
n subsystems 
ubsystem; there 
other pairs of 

between the non- 
n a loss of data 

system more 
For the system 
system must be 
of the other sub- 
ntral subsystem 
subsystems de- 
irs of the 
ectly . 

Analysis of the information content, as well 
as the potential demand and usage modes, suggests 
that most, if not all, of the proposed subsystems 
should be combinations of manual and automated 
systems, where the automation may be either me- 
chanical or electronic or both. There does not 
appear to be any justification, for example, for 
the extremely costly conversion of old drawings 
and maps to digital form; their value is primarily 
historical and the probability of use is relatively 
low for on-going park management purposes. How- 
ever, they should be cataloged so they are acces- 
sible through the system for scientific study 
(and perhaps legal) purposes. Manual and/or 
microform storage would seem appropriate for these 
types of informational materials. A similar case 
can be made for the bibliographic materials. 
Little justification exists for digitizing the 
text of books, reports, etc.; they should reside 
in libraries. The rare, one-of-a-kind documents 
should be microfilmed, however, to provide 
security backup. Specific recommendations for 
the manual and automated components of the in- 
dividual subsystems would be the subject of a more 
detailed design level of the study, assuming the 
proposed global design is adopted. 

The question of exactly which components 
would comprise the NPS EEMIS is not one easily 
answered. Certainly it seems clear that the 
Resources Data System (RDS) , which ideally fills 
the requirements for the central coordinating 
mode, is a proper subset of the EEMIS. It also 
seems desirable to include in the EEMIS those 
subsystems which process significant collections 
of data common to the RDS, regardless of whether 
it is preprocessing (i.e., serves as input to the 
RDS) or postprocessing (e.g., analysis). Inclu- 
sion of subsystems which serve primarily as 
input sources to the RDS for pointers, status 
items, or quantitative data (e.g., document 
identifiers or visitor counts) is optional, but 
it would seem desirable to include those sub- 
systems which have as their primary purpose 
support to the ecological and environmental mis- 
sions of the parks. Based on these guidelines, 
EEMIS should include, as a minimum, five basic 
subsystems: the RDS, Bibliographic System, 
Planning Management System, Graphics System, and 
Archaeology and Specimen System. Other subsystems 
which would lie outside the scope of the EEMIS 
under these guidelines, but which provide input 
to the RDS, are shown to provide perspective on 
the relationship of the EEMIS to still other NPS 
information systems. These other subsystems: 
Inventory and Maintenance System, visitor Re- 
servation System, and Law Enforcement System do 
not necessarily constitute a complete set, how- 
ever, as there will undoubtedly be others which 
should be included in both categories. Nor 
does the schematic representation attempt to 
show other existing NPS systems which have no 
direct input to or output from the central RDS 
system (e.g., financial and administrative 
systems) . 

The names of the proposed subsystems have 
been deliberately generalized to avoid any im- 
plication that they correspond to existing 
manual or computer-based systems. Rather, they 
are intended to identify the major types of data 
which are processed. 

The central s 
called the Natur 
be basically an 
the four major c 
Cultural , and So 
this central sub 
data called "poi 
user to informat 
(e.g. , the numbe 
etc.) Access to 
of the operation 
parks, regions, 
of data in the i 
trails, visitor 
item of data in 
subsystem just s 

ubsystem, the hub of the EEMIS, 
al Resources Data System, would 
inventory of data belonging to 
ategories: Physical, Biological, 
cioeconomic. Included also in 
system would be other types of 
nters" which would lead the 
ion located in other subsystems 
rs of drawings, maps, documents, 

this data is required in terms 
al units of the NPS (e.g., 
and WASO) and also by the types 
nventory (e.g., flora, fauna, 
use, etc.). In effect, every 
the master data base of this 
erves as an access point. 

The Bibliographic System provides the de- 
scriptive cataloging (e.g., titles, authors, 
and reference) , associated surrogate informa- 
tion (e..g., index entries, catalog codes, and 
abstracts), physical location, media charac- 
teristics, and ownership information for docu- 
ments of all types. These include but are not 
limited to books, technical reports, manuscripts 
and similar informational materials commonly 
found in libraries or personal reference collec- 
tions. It will also include information re- 
corded in non-print media (e.g., sound record- 
ings, slides, films, magnetic tapes, and micro- 
forms of various kinds.) If desired, it can 
also include references to resource people or 
facilities. Reports generated as products of 

2 2 

the planning functions 
statements, wilderness 
recorded in this system 
published. This subsys 
important components: 
ponent, which provides 
ments, and the collecti 
itself, which will rema 
(i.e., print or non-pri 

(e.g., environmental impact 
studies, etc.) would be 

as they were completed and 
tern consists of two equally 
the computer-based com- 
the access to the docu- 
on of information resources 
in in their present media 
nt) and locations. 

The Archaeologic 
vides the inventory 
archaeological and h 
The information cont 
elude a description 
istics (e.g., size, 
object, value, etc.) 
site, county, state, 
maintained for exhib 

al and Specimen Subsystem pro- 
and cataloging records for 
istorical items and specimens, 
ained in this system will in- 
of the object, its character- 
date found, probable date of 
, and its location (e.g., 
etc.). Records may also be 
ition purposes. 

The Graphics Subsystem is one of the largest 
and most complex of the proposed subsystems. 
First, it includes a component like that of the 
Bibliographic Subsystem for providing access to 
the various collections of graphics by such data 
as the name of the map or drawing, geographical 
coordinates, type (e.g., vegetation, hydrology, 
etc.), and similar characterizing information. 
It will also include descriptive information, 
such as the scale, source, physical size (for 
the print medium) , density and track if computer- 
readable, physical location, ownership and 
similar attributes. A second component of the 
Graphics Subsystem is, of course, the graphics 
materials themselves (i.e., the maps, drawings, 
photographies, etc.). As in the case of the bib- 
liographic resources, it is assumed these, too, 
will remain in their present media, both print 
and non-print, and locations. In the future it 
is expected that much of the graphics information 
will be obtained in computer-readable form from 
projects like ERTS and Skylab, where graphic 
information is normally stored in digital form 
on magnetic tape. The third major component, and 
the one which makes the Graphics Subsystem unique 
among those proposed, is the display component 
required for those graphics materials which are 
stored in machine-readable form. This component 
encompasses all of the processing required to 
accept as input digitized graphics data and to 
produce for use maps or overlays with the desired 
information at the appropriate scales. Thus, this 
component includes processing functions for in- 
tegrating ground truth data, adjusting for such 
conditions as curvature and atmospheric condi- 
tions, and plotting results for human use. 

The Planning Management Subsystem is primarily 
a task management system for control of planning 
projects. As such, it would contain information 
which follows a project from its inception to its 
completion, providing such management reporting 
assistance as schedules, task assignments, status 
reports, PERT or Gantt charts, and budgetary 
summarization. Its primary purposes are support 
for active planning projects and information for 
the "customers" (i.e., park personnel) or subcon- 
tractors as to overall project status. 

The Visitor Reservation Subsystem, which lies 
outside the scope of the EEMIS as defined, has 
as its primary function the scheduling or reser- 
vations for use of the various park facilities 
(e.g., camp sites, trails, wilderness areas, 
etc.) . Primary input to the RDS from this system 
is visitor usage data which is expected to be of 
use in modeling and simulation studies. 

The Law Enforcement Subsystem is basically an 
administrative reporting system through which 

historical records of various law enforcement 
activities are maintained, with periodic summary 
reports produced for management use. 

The Inventory and Maintenance Subsystem is 
intended as an inventory control system for all 
physical property and facilities, which includes 
buildings, trails, roads, signs, and similar 
facilities, regardless of cost. The principal 
function of this system is to provide support to 
the maintenance personnel of the NPS , including 
scheduled preventive maintenance, budgetary plan- 
ning, and historical records on these types of 
activities. Input to the RDS consists primarily 
of updates to the physical facilities records 
v/hen changes, additions, or dispositions are 
made, especially those related in any manner to 
the transportation system (including trails). It 
is also possible that boundary changes would be 
entered to the RDS via this subsystem. 

The planning stage for the EEMIS is now com- 
pleted. The move to make it operational is the 
next step. There are many exciting possibilities 
if we allow ourselves to think about what the 
future might hold. 

A park superintend 
minal on his desk would 
a network of national t 
he could consult a nati 
terized library; and he 
diately for answering q 
and modeling, and for f 
This superintendent wou 
selected maps, legislat 
compilations, technical 
specimens, computer sof 
of UPS staff — in short, 
formation concerning no 
all of the other park s 

ent with a computer ter- 

have immediate access to 
echnical information banks; 
onal environmental compu- 

could obtain data imme- 
uestions, for simulation 
iling necessary reports. 
Id have at his fingertips 
ion, manuscripts, data 

reports, books, films, 
tware, personal expertise 

a full range of NPS in- 
t only his park but also 
ites in the Service. 

It would be possible to determine the loca- 
tion of old maps, old drawings, and old reports 
that might be of use in decision-making. Infor- 
mation would also be immediately available con- 
cerning the location of non-print media such as 
sound recordings, slides, films, magnetic tapes, 
and microforms of various kinds. Generated re- 
ports such as environmental impact statements and 
wilderness studies as well as an inventory and 
cataloging record, for archeological and historic- 
al items and specimens would be immediately 
accessible. A superintendent would be able to 
follow any project from its inception to its com- 
pletion, using such management reporting assis- 
tance as schedules, task assignments, status 
reports, PERT or Gantt charts, and budgetary 
summarization. The superintendent and maintenance 
chief would have an inventory control system for 
all physical property and facilities, which would 
include buildings, trails, roads, signs and 
similar facilities. In fact, only human imagina- 
tion limits the possibilities. 

Certainly there are an 
the technology exists now. 
time, money, and strong mot 
prohibitively expensive, bo 
as well as probably technic 
accomplish all that has bee 
short period of time. It i 
therefore, to consider phas 
tion of subsystems one at a 
allow several things: (1) 
the NPS personnel both as o 
of the various components, 
of visible progress from th 
(3) prior investments in ex 

d will be problems, but 

What is needed is 
ivation. It would be 
th in time and money 
ally impossible to 
n envisioned in a 
s more realistic, 
ing in the implementa- 

time. This would 
time for education of 
perators and as users 
(2) an appearance 
e user's viewpoint, 
isting computer-based 


processing systems such as the graphic display 
modules developed for the Great Smoky Mountains 
RBI to be incorporated either on a temporary or 
permanent basis. In addition to a phasing in of 
the subsystems, it would be feasible to set up a 
few parks as prototypes, profiting from each ex- 
perience. The planning has been done and now the 
motivation has to come from the Park Service it- 
self, not from the Washington people, but from the 
level of those who have to deal with the parks on 
a day-to-day basis, those who would benefit most. 

J r > 



Richard G. Wiegert 



The National 
their varied ha 
tation and unna 
field research 
stone National 
warding, in par 
mostly because 
mal features . 
cal and biologi 
communities sha 
temperature, pr 
great outdoor 1 

Parks of our country, because of 
bitats and protection from exploi- 
tural perturbation, are ideal for 
in ecology. Of them all, Yellow- 
Park (YNP) is one of the most re- 
t because of size and age, but 
of its unique and abundant ther- 
Varying greatly in physical, chemi- 
cal characteristics, these thermal 
re the common thread of elevated 
oviding a unique experience and a 
aboratory . 

The most outstanding thermal features of YNP 
are its many geysers and boiling pools. However, 
the communities developing in the outflows of 
these thermal features tend, because of the high 
temperatures, to be very simple taxonomically , 
often consisting solely of bacteria, seldom sus- 
taining an animal grazing community and its atten- 
dant predators and parasites. Features such as the 
"acid outflows" support organisms that produce 
sulfuric acid and the resultant low pH attracts 
unique communities of organisms. With both high 
temperature and low pH , these communities also 
comprise only a few species of microorganisms . 
At lower temperatures very interesting animal 
communities develop in these acid systems, but 
the pH is very difficult to maintain in the 
laboratory and ecological work with these systems 
is hampered. 

In the so-called "alkaline" spring areas, 
principally in the band of thermal features called 
Upper, Midway, and Lower Geyser Basins, condi- 
tions of temperature and pH are found which are 
neither too high nor too low, but are, as Goldi- 
locks found the porridge and the bed of Baby Bear, 
"just right", for the development of an extremely 
interesting community of algae, arthropods, and 
bacteria . 

In a sense I was preadapted when, in the summer 
of 1967, Tom and Louise Brock invited me to visit 
their Yellowstone laboratory and spend a few days 
doing radio isotope studies of the grazing food 
chain in these alkaline thermal communities . I 
was just then interested in the budding discipline 
of systems ecology but was becoming discouraged by 
the difficulties of beginning with models of the 
usual complex communities of old field, lake, or 
forest. I had for the past two years been look- 
ing for a naturally occurring community that was 

1) Simple in the sense of containing few species 
yet containing all the usual ecological inter- 
actions (grazing, decomposition, predation, etc.), 

2) Easy to sample, 3) Amenable to manipulation in 
the field, 4) Subject to laboratory study, at 
least at the level of the individual populations 
and 5) Capable of rapid successional change and 
the attainment of climax or steady state in weeks 
or months rather than decades or centuries. In 
short, I wanted an ecosystem where a small group 

Department of Zoology, University of Georgia, 
Athens 30602. 

of individuals could, in a few years, study in 
considerable detail the biology of at least all 
major groups of organisms and in many cases 
examine them species by species . The objective 
would be to construct a computer model of the 
system that would not only reproduce the normal 
states of the community, but would have the 
capability of predicting the behavior of the 
systems under a variety of perturbations. This 
model could in turn provide some insight into the 
kind of procedures that might be employed for the 
construction of models of the more complex systems 
on which the economic and recreational well being 
of man depend. 

In 1967, on my first visit to the Brock's 
study area, I realized that I had found what I was 
looking for. For the past 8 years the ecology of 
the thermal algae-brine fly communities of YNP has 
been a major interest of mine. In this paper I 
wish to discuss briefly the major interactions 
and, by so doing, to tell the story of the de- 
velopment of an ecosystem model, a story still un- 
finished, but whose end is now known, if not yet 
written down. 

This has truly been a team research effort, 
and I am indebted to all of my colleagues, in- 
cluding the five Ph.D. students, eight post- 
doctoral associates and 14 technicians who have 
been directly involved with this research over the 
past 8 years. The research has been supported by 
NSF Grants GB 21255 and GB 76813. 


In broad outline the story of successional life 
and death in an alkaline thermal-spring community 
is simple and straightforward. Within a few 
weeks of beginning our studies in 1968 the follow- 
ing sequence was worked out. At temperatures 
between about 40-50°C a mat of filamentous blue- 
green algae begins to grow on any firm substrate. 
The growth of this algal mat eventually changes 
what was originally a smooth sheet of water into 
a series of channelled flows. With continued 
growth of the algae, parts of the mat eventually 
reach the surface and cool, thus producing a 
heterogeneous mosaic of temperatures. The algae 
in these cool areas of the mat are no longer 
actively growing, for the nutrients supplied by 
the flowing water are cut off, but the tempera- 
ture of these areas is now low enough (<40°C) so 
that the grazing brine flies (Ephydridae) can 
survive and they quickly saturate the area with 
eggs. The larvae of these flies begin to feed 
on the algal-bacterial mat, in the process 
breaking down its structure and "solubilizing" 
it until at the end the larvae are moving about 
in a semiliquid medium of water and fecal pel- 
lets, the latter often containing many algal 
cells and parts of filaments, as shown by U.V. 
microscopy . 

Eventually, the larvae (many of which pupated 
and left as adults) eat through some critical 


algal "dam" and hot water pours into the cool 
patch killing the remaining larvae, washing out 
the remaining cellular debris and starting the 
process of algal growth all over again. Thus at 
any point in space the community is in a constant 
state of flux, but over any larger area of algal 
mat, the standing crops of both algae and flies 
may exhibit a steady state. 


Constructing a model of an ecosystem involves 
a number of preliminary steps of the sort re- 
quired for any ecological study. First an inven- 
tory of the species or at least the major func- 
tional groups of organisms must be drawn up. Next, 
their trophic interactions (and their behavioral in- 
teractions) must be worked out. Next, and most 
difficult of all, the functional form of the factors 
controlling the rate at which each group processes 
material of energy must be elucidated. Initially 
this is verbalized, then transformed into a mathemat- 
ical expression and finally into a computer program 
whence simulation of community dynamics is possible. 

First working out the developmental sequence 
in the alkaline thermal community, our immediate 
task was to provide a species list and work out 
the food web. Naturally, all these steps are not 
in reality done in a nice tight sequence, but pro- 
ceed as larger saltations, returning to clean up 
problems bypassed for the convenience of the mo- 
ment. Nevertheless, eventually the task of cen- 
susing was completed, showing a species complement 
of approximately 8-10 species of algae, perhaps 
twice that many 'kinds' of bacterium, 3 species 
of ephydrid fly and several predators and para- 
sites, most of which reproduce and spend much or 
most of their life outside of the thermal algal 
mat proper . 

In the particular springs chosen for this work, 
the dominant producer species is the blue-green 
alga Mastigocladus laminosus . This is fed upon 
by the major grazer, a species of brine fly 
(Paracoenia turbida; Ephydridae; . The brine fly 
is in turn utilized by a long-legged predatory 
fly ( Tachytrechus angus ti penni s ; DolichopodidaeJ 
and a parasitic mite (Partnuniella thermalis ; 
Hydrachnellae; . Under certain conditions a pupal 
parasite, the wasp Urolepis rufipes (Pteromalidae) 
can be important in this community. A number of 
other predators remove flies, among these are the 
sandpiper, tiger beetle and wolf spider. These major 
trophic interactions form the basis for the model- 
ing efforts. The next step was to study, in detail, 
the biology of each group with a view toward re- 
vealing how it interacted with the various factors 
regulating growth and thus determining levels of 
energy flow and mineral cycling. As examples, I 
will discuss briefly the studies on algae, brine 
fly, and on long-legged fly, groups regulated pri- 
marily by an inorganic resource, intra-specif ic 
competition for space and behavioral interactions 

The Study Sites 

Study of naturally-occurring springs has serious 
limitations. For one thing, control of flow is 
difficult or impossible. Indeed, many natural 
springs may abruptly shift their outlets or fluc- 
tuate widely in rate of flow. Furthermore, large 
animals stepping into the channel constitute an 
ever present (and quite common) hazard. Further- 
more, the soft bottom of many natural algal sys- 
tems made sampling difficult; up to a foot or more 
of silt may accumulate over the hard silica or 
"sinter" that forms the base of the system. 

Thus the decision 
studies on communitie 
(called Boards) with 
end. These are all ] 
adjacent to Firehole 
Basin, and have been 
publications and rang 
meters in width and 1 
flow rates of up to 5 

was made early to conduct the 
s developed on wooden troughs 
water piped onto the upper 
ocated at Serendipity Meadow 
Lake Drive, Lower Geyser 
described in a number of 
e in size from a few centi- 
ength to 1.2 x 24 meters with 

A short paper cannot even summarize the results 
of 8 years of investigations by our team of re- 
searchers, therefore I have provided (in an ap- 
pendix) a list of the publications that have re- 
sulted from the work by the University of Georgia 
group in YNP . The three studies that are sum- 
marized here, namely those on: 1) producers, 
(blue-green algae), 2) the grazer, (Paracoenia 
turbida) and the predator (Tachytrechus angusti- 
pennis) were selected both for their importance 
in the system and to illustrate the differences 
in emphasis and procedure that must be employed 
to derive model-relevant parameters for organisms 
that differ greatly in the mechanisms controlling 
their utilization of resources. 



The filamentous blue-green algae characteristic 
of the thermal alkaline area waters in the tem- 
perature range 40-56 C are limited primarily by 
physical constraints plus intra and interspecific 
competition for a limited renewable material re- 
source, free carbon dioxide. Briefly, the initial 
observations were that under a given regimen of 
light intensity, day length temperature and flow 
conditions, the rate of net productivity decreased 
with distance from the source. A considerably 
smaller downstream decrease in the steady state 
standing crop and a shift in species dominance 
were also noted. Studies of photosynthesis rate, 
biomass and chlorophyll concentration showed a 
strong nonlinear relationship between these three 
variables . Gross photosynthesis was related to 
the 0.29 power of ash-free dry weight. Light in- 
tensity (at all levels above approximately 10% 
of full sunlight) showed no relationship to rate 
of gross photosynthesis. Respiration was linearly 
related to biomass; thus net rate of photosyn- 
thesis also followed the 0.29 power of biomass 
relationship . 

Because the photosynthesis measurements were 
carried out in bottles under identical conditions 
of nutrient availability (source water from the 
single spring was used in all experiments) we next 
looked for any factor on the experimental board 
troughs that could have been correlated with the 
downstream decrease in rate of net productivity 
(the distance and volume of flow were such that 
temperature decrease downstream averaged less 
than 2°C) . The only factor of this kind that 
could be identified was the concentration of free 
CO ? . From a high of 95 ppm, C0 2 declined to 15 
ppm on the last of four short (2m) wooden troughs. 
The hypothesis was that the high levels of C0 2 in 
the source water were stimulating the high produc- 
tivity near the source because the thermal water 
contained very high quantities of available 
phosphorus and the algae themselves are nitrogen 
fixers . The only way the available COy can reach 
the vicinity of the individual filament is by 
diffusion through the interstitial water of the 
mat, although studies with micro pH probes showed 
that the first 1-2 mm of mat may be supplied 
with nutrients by direct replacement of water 
through turbulent flow. Putting all this infor- 
mation into the form of a model of algal growth 
the predictions of the model provided an 


acceptable explanation of the downstream decrease 
in rate of net production and an approximation of 
the measured steady state. 



The brine fly Paracoenia turbida is an animal 
approximately one half as large as a house fly 
with a surprisingly large egg production and 
growth rate, giving a potential for increase that 
is the equivalent of that of Drosophila . 

0.25 flies X fly" day 


From the mo 
rate of increa 
corrected by d 
that is assimi 
maximum specif 
lation could i 
ingestion has 
multiplied by 
the instantane 
units of food 

deling standpoint, this potential 
se is a vital parameter for, when 
ividing by the fraction of food 
lated, it becomes an estimate of the 
ic rate at which a growing fly popu- 
ngest food. This specific rate of 
the dimensions time - 1 and, when 
the standing crop of flies, gives 
ous rate of ingestion measure in 
per unit time. 

The life history of the brine fly Paracoenia 
is not complex. The adults feed from an oviposit 
on the algal mat. The eggs hatch in two or more 
days (optimum development temperature is 35°C) . 
The larvae pass through 3 instars in the course 
of approximately 6 days, pupate, and emerge 
as adults some 6 days later. The total life 
cycle from egg to egg can occur in a minimum of 
14 days. 

Behavioral factors have little importance in 
controlling the numbers of flies . The adults 
feed on the surface of the algal mat and always 
have an abundance of food. Studies of ovi- 
position showed no effect of crowding up to the 
point where the flies became so dense that some 
could not reach the surface of the algae. How- 
ever, because the eggs and larvae are killed by 
temperatures only slightly greater than 40 °C, 
the only available habitat for the immature forms 
is in temporary cool spots in the algal mat 
caused by the growth of the algal shifting of hot 
water. Dense masses of eggs are deposited in 
these temporary cooler areas and the larvae 
quickly eat up most of the mat and destroy the 
filamentous structure. As more and more larvae 
enter the area (by continued egg deposition) some 
of the oldest larvae pupate and leave as adults. 
The remaining larvae soon show the effects of 
crowding; the growth rate declines, time to pupa- 
tion grows longer, pupation occurs at a smaller 
average age, and mortality of both larvae and 
pupae increases. Studies of the crowding effect 
on larvae showed a threshold effect, probably due 
to the buildup of toxic waste products in the pres- 
ence of abundant food. Although under laboratory 
conditions, the larvae can consume all their algal 
food and subsequently starve, in the field the 
crowding effect invariably begins long before the 
algae are gone because the cool patches have 
little or no water flowing through them, thus 
allowing toxic wastes to accumulate. Alternatively, 
the larvae may break down the algal mat holding 
back the hot water, causing their death and the 
death of many of the pupae. 

Field and laboratory research on this species 
was directed toward the measurement of the various 
rates of inqestion and loss as well as the thresh- 
old of standing crop of flies and algae that 
determine whether flies are in an optimum en- 
vironment or not. The result was definition of a 
community in which three distinct conditions can 
be identified: 1) a hot-flowing condition from 

which flies are excluded and algal growth is 
occurring, 2) a hot-stagnant condition with no 
net algal growth and from which fly eggs and 
larvae only are excluded, 3) a cool-stagnant 
condition in which both adult and larval flies 
are decimating the algae. At a given point there- 
fore, the system is never constant; it is either 
growing or being destroyed. Over several square 
meters, however, these communities do often show 
considerable constancy of algal biomass and fly 
density. Superimposed on this pattern of fly- 
algae interactions is predation (and parasitism) 
of the brine flies. Except for an occasional ex- 
ceptional circumstance of extreme fly abundance 
and low algal standing crop, the predators and 
parasites in this system do not control the flies; 
instead they are often themselves controlled by 
the long term fluctuations in fly density. 



A prominent member of the predator community 
is the long-legged fly, (Dolichopodidae) 
Tachytrechus angustipennis . The adult is the only 
life history stage of this species on the algal 
mats. The female fly lays its eggs in the fall 
(near the end of its life) and the larvae and 
pupae exist in the mud at the edges of the thermal 
effluents . 

The adult is an active predator which seeks out 
the eggs and small larvae of Paracoenia . In con- 
trast to the latter fly, behavior is an important 
factor controlling the abundance and predatory im- 
pact of Tachytrechus . Thus, a large share of the 
time spent studying this predator was devoted to 
observation. Tachytrechus males establish many 
short-lived territories during a day, from which 
they attempt to exclude all other insects except 
females of their species. Although the primary 
function of this behavior may be reproductive, it 
does result in a limit placed on the number of 
dolichopodids that can search a given area at the 
same time. 

These flies are vey inefficient at extracting 
eggs of Paracoenia from the algal mat because 
these eggs are covered with recurved hooks which 
hold the eggs firmly in place. Larvae are some- 
what more available, but the predator has diffi- 
culty handling anything larger than about 2nd 
instar. The major increase in size of the larva 
takes place during the 3 instar (half the total 
estimated development time) , thus under normal 
circumstances most of the larval standing crop is 
unavailable to the predator. This fact, together 
with the single annual generation and low biotic 
potential of the dolichopodid fly renders it in- 
capable of making serious inroads on its prey 
population . 


Information f 
described was us 
dictive computer 
getics of the th 
community. Mode 
forming experime 
the field. For 
of water to crea 
mat, thus increa 
fly density. 

rom studies such as those just 
ed to construct detailed pre- 

models of the ecological ener- 
ermal algae consumer-predator 
1 predictions were tested by per- 
nts on the entire ecosystem in 
example, shutting down the flow 
te larger areas of cool-stagnant 
sing, at least temporarily, the 

Once the most detailed model was tested suf- 
ficiently to establish its validity under a lim- 
ited range of conditions, the model itself was 
rearranged in different configurations, each 
simpler than the last. For example, instead of 


keeping track of each separate life history stage 
of the brine fly, various condensations were made 
ranging from all larval stages in one compartment 
to all brine flies in one compartment. Instead of 
mimicking the normal time delays in passage from 
egg to larva to pupa, the time delays were only 
approximated or removed entirely, etc. These 
condensed versions of the model were used to simu- 
late the behavior of the system under a range of 
perturbations. These simulations were then ex- 
amined to determine the degree to which a particu- 
lar type of condensation had affected the pre- 
dictions of the most detailed model. Out of this 
work, hopefully, will come some rules for model- 
ing ecosystems that will prove useful to those 
contemplating models of larger complex systems 
that are vital to the well-being of man. The 
early progress appears promising and the work on 
the Yellowstone thermal communities is continuing. 


1969. Feedings by Paracoenia and Ephedra 
(Diptera: Ephydridae) on the micro-organisms 
of hot springs. Ecology 50:192-200. (Appen- 
dix II) . 

COLLINS, N. C. Mechanisms determining the rela- 
tive abundance of three brine flies (Diptera: 
Ephydridae) in thermal spring effluents, 
(submitted to Canadian Entomologist) . 

COLLINS, N. C. 1975. Population biology of a 
brine fly (Diptera: Ephydridae) in the pres- 
ence of abundant algal food. Ecology 56: 

COLLINS, N. C. 1975. Tactics of host exploita- 
tion by a thermophilic water mite. Misc. Publ. 
Entomol. Soc. Am. 9:250-254. 

Functional analysis of a thermal spring eco- 
system, with an evaluation of the role of con- 
sumers. Ecology (in press). 

FRALEIGH, P. C. and R. G. WIEGERT. 1975. A 

model explaining successional change in stand- 
ing crop of thermal blue-green algae. Ecology 

GORDEN, R. W. and R. G. WIEGERT. Bacterial types 
and interactions in a thermal blue-green algae- 
ephydrid fly ecosystem. in: Aquatic Microbial 
Communities, John Cairns, (ed) . (in press). 

KUENZEL, W. J. and R. G. WIEGERT. 1973. Ener- 
getics of a Spotted Sandpiper feeding on brine 
fly larvae (Paracoenia ; Diptera; Ephydridae) 
in a thermal spring community. Wilson Bull. 
December . 

KUENZEL, W. J. and R. G. WIEGERT. Yellowstone 
thermal effluent systems: Life cycle popula- 
tion dynamics, territoriality and energetics 
of a major insect predator (Diptera, Dolicho- 
podidae. Oikos (in press). 

MITCHELL, R. 1974. The evolution of thermophily 
in hot springs. Quart. Rev. Biol. 49:229-242. 

MITCHELL, R. and B. L. REDMOND. 1974. Fine 
structure and respiration of the eggs of two 
ephydrid flies (Diptera: Ephydridae). Trans. 
Am. Micros. Soc. 93 (1 ): 113-118 . 

OWEN, D. F. and R. G. WIEGERT. 1976. Do consum- 
ers maximize plant fitness? Oikos 27:3(1-5). 

WIEGERT, R. G. and P. C. FRALEIGH. 1972. Ecolo- 
gy of Yellowstone thermal effluent systems: 
net primary production and species diversity 
of a successional blue-green algal mat. Lim. 
& Ocean. 17:215-228. 

WIEGERT, R. G. and R. MITCHELL. 1973. Ecology of 
Yellowstone thermal effluent systems: Inter- 
sects of blue-green algae, grazing flies 
(Paracoenia , Ephydridae) and water mites 
(Partnuniella , Hydrachnellae) Hydrobiologia 

WIEGERT, R. g. 1974. A general ecological model 
and its use in simulating algal-fly energetics 
in a thermal spring community. in: P. W. 
Geier, L. R. Clark, D. J. Anderson and H. A. 
Nix (Eds.) Insects: Studies in population 
management. Vol 1, pp. 85-102. Occasional 
Papers. Ecol. Soc. Austr., Canberra. 

WIEGERT, R. G. 1974 Competition, a theory based 
on realistic general equations of population 
growth. Science 185:539-542. 

WIEGERT, R. G. 1974. A general mathematical 
representation of ecological flux processes: 
description and use in ecosystem models. 
Proc . 6th Ann. S. E. Systems Sypm., R. Kinney 
(Ed.) TF-2 15 p. Baton Rouqe . La. 

WIEGERT, R. G. 1975. Mathematical Representa- 
tion of ecological interactions. in: Ecosys- 
tem Analysis and Prediction. Proc. SIAM-SIMS 
Conf., Alta, Utah, S. A. Levin, ed . Philadel- 
phia Soc. Ind. and Appl . Math. 

WIEGERT, R. G. 1975. Simulation modeling of the 
algal-fly components of a thermal ecosystem: 
Effects of spatial heterogeneity, time delays 
and model condensation. in: Systems Analyses 
and Simulations in Ecology. B. C. Patten Ed. 
Vol. 3, Academic Press, N. Y. pp. 157-181. 

WIEGERT, R. G. 1975. Simulation models of eco- 
systems. Ann. Rev. Ecol. and Syst. 6:311-338. 

WIEGERT, R. G. 1977. Population models: experi- 
mental tools for analysis of ecosystems. Pro- 
ceedings of Colloquium on Analysis of Ecosys- 
tems. Ohio State University Press, Columbus. 
(In Press) . B. J. Horn, R. Mitchell and G. R. 
Stairs, Eds. 

WIEGERT, R. G. 1977. in: Models as Ecological 
Tools: Theory & Case Histories. (In Press) . 
C. Hall and J. Day, Editors. J. Wiley and 
Sons . 


Edward DeBellevue, Howard T. Odum, Joan Browder and George Gardner 2 


The design of humanity and nature now develop- 
ing over the American landscape consists of a 
mixture of farms, urban areas, and natural areas 
and parks set aside for their special values. 
These contribute to the total economy of man and 
nature in many ways, some well-known and some 
under-evaluated. In applying systems approaches 
to understand the larger system of humanity and 
nature, many questions arise about the interac- 
tions of the parks with the other components of 
the environment. How much park area is needed in 
ratio to developed areas? How much do economic 
use and interfacing development threaten park 
values? How do park values compare with trad- 
itional economic values of developed areas? 

Energy models may provide a 
and portray a park in intermed 
enough detail to show major in 
in so much detail that one can 
tion. After pathways are eval 
energy analysis suggests answe 
about comparative values. In 
analysis is made of the Evergl 
and suggestions are made about 

method to evaluate 
iate complexity with 
teractions, but not 
not visualize opera- 
uated numerically, 
rs to questions 
this paper an energy 
ades National Park 

impact interactions. 

A national park is a complex environmental 
system which includes many component ecosystems, a 
pattern of management and human visitation, and, 
externally, a set of energy flows and economic 
interactions between the park and the economy of 
the county, the state, and the nation. This is a 
first attempt to show a national park as an in- 
tegral part of a larger system and to point out 
what may be the most important components and 
interactions with model diagrams. 

This is a component of the South Florida Study, 
Phase II of the Department of the Interior. A 
summary report has just been printed (Browder, 

We thank the following agencies and persons 
for their time and efforts helping us obtain data 
and understanding used in evaluating the Ever- 
glades National Park Model: National Park 
Service, Washington, D.C.: J. Matthews; Ever- 
glades National Park Office: Assistant Superin- 
tendent Claude McClain, E. Countreman, G. Davis, 
T. Schmidt, Larry Bancroft, Frank Nix, Jerry 
Hammond, Ralph Mealey, Roger Ronek, Glen Garrer, 
and numerous other Park staff; National Marine 
Fisheries, Miami: J. Cato; National Audubon 
Society, Inc., Corkscrew Swamp Sanctuary: M. 
Duever; Central and Southern Flood Control Dis- 
trict, WEst Palm Beath: W. Brannon; Florida 
State Board of Health, Vero Beach: G. Omear; U. 
S. Army Corps of Engineers, Jacksonville: G. 
Beque; and Homestead Chamber of Commerce: D. 
Cappisalo. This work was accomplished under con- 
tract with the Department of the Interior, 
National Park Service, #PX-0001-60080. 

Center for Wetlands, University of Florida, 

Gainesville 32611. 

Littlejohn, and Young, 1976). The large technical 
report which has details on the other systems of 
region, county, and ecological communities is 
being revised for publication. 


Ideas and concepts of Everglades National Park 
in terms of its main components and pathways of 
interaction and influence were sought from Staff 
of Everglades Park and other sources and used to 
develop a preliminary model, in diagram form, 
using energy circuit symbols (Fig. 1). Discus- 
sion of the energy circuit language and the 
energetics approach to systems analysis can be 
found in two books (Odum, 1971; and Odum and 
Odum, 1976) . 

Following the development of the Park model, 
stocks and pathways were evaluated with available 
data. Stocks and flows were converted to equiva- 
lent energy values in coal equivalents based on 
the energy cost of their replacement. These data 
sources and calculations are presented in tabu- 
lar form and summary numbers have been placed on 
appropriate figures. Quantification of stocks 
and flows was aided by an unpublished technical 
report of the South Florida Study, Phase II, in 
which were presented subsystem models for sys- 
tems found within the Park such as mangrove, 
saltwater marsh, sawgrass, and wet prairie eco- 
systems, etc. Major flows and storages were 
aggregated to show perspective on major areas of 
interest, such as total income, total environ- 
mental work done by the Park's natural systems, 
total energy, and energy based purchases related 
to the Park. 

Due to lack of available information on money 
spent outside of the Park by visitors to the 
Park, an indirect calculation was made by mul- 
tiplying the number of visitors to the Park by 
days in the Park and by the normal tourist ex- 
penditure per day in Dade County. For each day 
in the Park, it was assumed that the person 
extended his stay in Florida by a day and, in 
the process of entering and leaving, spent this 
amount. Unfortunately, it was not possible to 
separate tourists from local visitors in the 
information provided by the Park, so an estimate 
was made that 10 percent of the visitors were 
local. The resulting figure is probably an 
underestimate because the number of tourists 
other than fishermen entering the Park from 
Everglades City, Chokoloskee, and the Florida 
Keys could not be counted. 

The economic value of the Park fishery in 
terms of both commercial and sports fishing was 
estimated, based on the reported kilograms per 
year commercial and sport fish catch at Flamingo. 
Both catches were multiplied by the retail price 
of the fish as an estimate of their total value 
to the economy. The annual retail value of the 
Dry Tortugas shrimp fishery also was counted. 

Use of the energy symbol language in model 
diagrams carries with it implications as to the 
mathematical relationships and performance of 


the system with time. Two recent publications 
from the South Florida Study illustrate and fur- 
ther describe the energy circuit diagramming, 
evaluations, and simulation results (Zuchetto, 
1976; and Bayley and Odum, 1976) . 

Calculation of Ratio of Work of the Economy 
to Work of the Park's Natural Ecosystems 

The model was further aggregated by combining 
flows into economic and natural categories and a 
ratio was calculated between the work of the asso- 
ciated economy and the work of the Park ecosystems 
(both in units of coal equivalent kilocalories) . 
This ratio of fossil fuel energy input to natural 
energy input is called the investment ratio. The 
Park systems' investment ratio was compared to 
that of the United States, 2.5, to give perspec- 
tive about the impact of the park on the economic 
system and vice versa. 

Expressing Data in Coal Equivalents 

Table 1 shows the conversion factors used to 
estimate the energetic value of the stocks and 
flows evaluated for the model. The common energy 
unit is the "coal equivalent kilocalorie" 
(CE kcal) . Footnotes to the table give the basis 
for the conversion factors, determined by means 
of analyzing systems in which the conversion of 
one kind of energy to another, such as coal to 
electricity, has been evaluated, showing all 
energy inputs and subsidies. The kilocalorie/ 
dollar conversion was obtained by dividing the sum 
of the annual U.S. fuel usage and annual solar 
radiation (in CE kcals) by the gross national 
product (GNP) ; this ratio represents the energy 
invested for every dollar circulating in the na- 
tional economy. Values understood in terms of 
dollars, weight, speed, etc., can be expressed 
in coal equivalents by use of the appropriate con- 
version factors. In most cases, resultant values 
represent replacement cost (how much energy that 
would have to be expended to recreate) . See the 
Appendix for further discussion of the conversion 

Application of Decision Making 

The use of energy conversion factors and the 
investment ratio to aid decision-making was demon- 
strated with an exercise quantifying the potential 
impact of an increased water supply for the Park 
on the Park's value to the regional system. 


Results include a land use map of ecosystems in 
the Park, an evaluated model of the greater Park 
system, and summary diagrams of exchanges. 

Distribution of Ecosystems in the Park 

Given in Fig. 2 is the land use map of the 
Everglades National Park in 1973. From this map, 
areas of ecosystems were measured (Table 2) . Per- 
centages of major ecosystems are diagrammed in 
Fig. 3. The area of broad shallow marshland that 
gave the park its name represents only 9 percent 
of the Park's 1.4 million acres. An additional 
16 percent in wet prairie combines to make one 
quarter of the Park a freshwater wetlands. Inter- 
spersed are subtropical hammocks and pinelands 
which, although important in increasing the diver- 
sity and beauty of the Park, represent less than 
2 percent of the total area. About three quarters 
of the Park are estuarine ecosystems; one quarter 
mangrove, one quarter marine meadows, 7 percent 
salt marsh, and an additional 14 percent where 

plankton is the major primary producer. About 40 
percent of the Park is in Florida Bay, which in- 
cludes the marine meadows and planktonic systems. 
Thus , the Park is predominantly an estuarine 
park, although a visitor driving along its main 
access road (SR 20) would see about half wet 
prairie and half mangroves. Creation of the Park 
preserved only a small percent of the historic 
"river of grass." 

The Park's na 
three main categ 
water wetlands , 
responsible for 
of the Park's to 
tivity of approx 
Mangroves repres 
prairies account 
meadows of Flori 
the natural ener 
consisting of sa 
account for over 
productivity, pr 
they cover. 

tural systems can 
ories: freshwate 
and open water 
27, 39, and 33 pe 
tal annual gross 
imately 9.07 tril 
ent 33 percent of 

for 15 percent, 
da Bay account fo 
gy budget. The e 
ltwater wetlands 

two thirds of th 
imarily because o 

be grouped into 
r wetlands, salt- 
These systems are 
rcent respectively 
primary produc- 
lion CE kcal. 

the total. Wet 
and the marine 
r 29 percent of 
stuarine systems, 
and open water, 
e gross primary 
f the large area 

can be 
coal e 
are th 

ss primary producti 

used as a measure 
terns" of potential 
quivalent units (CE 
ratio of the total 
an estimate of $41 
tic value of the Pa 
ant pathways by whi 
e tourist industry 
fishing industries 
le and more difficu 
importance to the r 

vity (GPP) of ecosystems 
of the "work of natural 
benefit to man. GPP in 
kcal) multiplied by the 
economy, $19,600 kcal/$, 
2 million/yr. for the 
rk to the region. Two 
ch this value is realized 
and the commercial and 

Other pathways less 
It to quantify may be of 
egion's economic well- 

The Evaluated Model 

The model of Everglades National Park is shown 
in Fig. 4, and values are given in Table 3. The 
calculations and sources leading to the quanti- 
ties in Table 3 are given in Appendix Table Al. 
On the left hand side of Fig. 4a and in greater 
detail in Fig. 4b are shown the interactions of 
sun, wind, rain, and tide; which produce the ham- 
mocks, and pinelands, sloughs and prairies, man- 
groves and salt marshes, and the plankton and 
submerged meadows of Florida Bay. From the energy 
flows, expressed as gross primary productivity, 
are generated the wilderness and wildlife values 
that attract visitors, produce food fish and shell- 
fish, cause economic investments outside the Park, 
and even give esthetic value for people who never 
visit but appreciate it (right hand side of 
Fig. 4a) . 

Economic flows respond to the pumping action of 
this intangible quality termed "image." The 
Park's image may ultimately be dependant on the 
balance between production of value from all 
sources and loss of value with deaths, fading 
memory, and negative reputation factors, such as 
the abundance of summer mosquitoes or intrusions 
from adjacent urban or agricultural development. 
Water is an important input to the productivity 
and character of nature ecosystems and one which 
has greatly been changed by man's drainage works 
in the upper Everglades basin. 

Inflows and Outflows 

Identified and evaluated in Fig. 5 are the 
major driving energy sources and flows in the 
form of matter, dollars and information, which 
flow between Everglades National Park and the 

J 2 

Energy source (forcing function), 
source of external cause. 

Heat sink, outflow of used energy. 

Energy Interaction, one f ype of energy 
amplifies energy of a different quality 
(usually a multiplier). 

Economic transaction and price functla 

Storage (state variable). 

Circulating energy transformer with 
Michaelie-Menton kinetics (diminishing 
returns transfer function) . 

On-off control work d'lgltal ictlonfi). 


Croup symbols (1) juror a t -i I y t U self- 
maintenance units, (2) production units, 
and (3) general purpose bo* for miscel- 
laneous subsystems. 

FIGURE 1. Symbols used in energy analysis diagrams, 

TABLE 1. Energy quality ratios used in this study. 

Energy Type 

Fossil Fuel Equivalence Factor 
(kcal FFe per kcal) 



Sugar of gross production 
still distributed over land 

Wood still distributed 
over land 

Winds (10 mph) 


Fossil Fuel (coal) 


Water Elevation Potential 

Water Chemical Potential 

Dollar Flow, 1970 

7 x 10 







By definition 



1. 7 

• i 


Appendix A 


Kemp (1977) 



25,000 (kcal 

FFE/S) a 

*Fossil fuel mined and stored at place of use 

a From Odum and Brown (eds.) (1975); and Odum et al. (1976) 


\ • -. ■■:■■ *-* "~-t'- *" ' •/ ' 1 

\ \* ^- ' --- . i' J — -'^■■■r.-'i f'» 


/ / r^-> 

- » » ■ ■ V ! *2ii> ■ '*«* a ■'Jk- / 

\ e.^ 





FIGURE 2. Ecosystems in Everglades National Park. 

13 6 %■ 

I 6% 


(a) LAND 







/--\ ^-WET PRAIR 
/ ><^ 15 0% 




/ 12 0% 





FIGURE 3. (a) Percent land areas of Everglades National Park by ecosystem type based on 
values from Table 2. (b) Percent gross primary production contribution by ecosystem type. 
Note: "Other" land use category not included. 


TABLE 2. Area of energetic subsystems of Everglades National Park for 1973. 

Name Map 




of area 
(%) 3 




(medium density) 





Commercial & 






Vegetable Crops 





Pineland Sys. 





Hardwood Sys. 






©omes & Strands) 





Wet Prairie 





Scrub Cypress 





Marshes & Sloughs 





Sawgrass Marsh 





Beach & Dunes 





Saline Transition 





Scrub Mangrove 





Saltwater Marsh 










Coastal Plankton 





Natural Embayments 





Medium Salinity 

Plankton Estuary 





Marine Meadows 








Source: measured from S. Fla. regional map by R. Costanza and T. Brown in 
Browder, Littlejohn, and Young, 1976. 


Rounded to nearest tenth percent, ten acres, ten hectares. 


FIGURE 4a. Systems model. Everglades National Park. 
TABLE 3. Values for model of the Everglades National Park and its external relationships (Fig. 4). : 

Footnote // 



1 Image 

2 Visitors 

3 Park capital assets 

Park related assets (outside park) 

5 Fishing related assets (boats, etc.) 

6 Water control capital assets 

7 Area of hardwoods & pinelands (incl tree islands) 

8 Area of marshes, sloughs & wet prairies 

9 Area of mangroves & s.w. marshes 

10 Area of bay 

11 Phosphorus in hammocks and pinelands 

12 Phosphorus in marshes, sloughs and wet prairies 

13 Phosphorus in mangroves 

14 Phosphorus in bay 

15 Groundwater 

16 Water in sloughs & wet prairies 

17 Water in mangroves 

18 Biomass of plants in hammocks & pinewoods 

19 Biomass of plants in slough and wet prairie 

20 Fiomass of trees in mangroves 

21 Biomass 

22 Organic matter in mangroves 

23 Wildlife, crayfish, fish & wading birds 

24 Mosquitoes in mangroves only 

2787 Visotors 


125 million $ 

$16.84xl0 6 

$2.11xl0 6 

23,000 acres 

351,640 acres 

465,310 acres 

547,570 acres 

0.369x10 kg 

738xl0 6 kg 

346xl0 6 kg 
0.353x10 kg 

1.85 million acre ft 

314x10 acre ft 

502x10 acre ft 

( * 

8.4x10 kg 

* ft 

980x1010 kg 

1659x10 kg 

407xl0 6 kg 



0.003x10 kg 

0.0081x10 kg 


in 9 

10 mos 


Footnote it 

: r i 



Invertebrates in bay ** 28.4x10 kg 


Water supply to north which is released to the park 266,000 ac _ it/y r 

26 Fish in bay 1.6x10 kg 

27 Terrestrial Wildlife no data available 

30 Availability of fuel 34.95x10 kcal/yr 

31 Electric capacity 38.68x10 kcal/yr 

32 Availability of goods & services 2.175x10 $/yr 

54.4x10 kcal/yr 

33 Availability of FCD monies $10,964,000 

34 Availability of Federal money 403,900,000 $/yr 


35 Population available for visiting 211.4x10 


36 Dollar availability to visiting population 977.3x10 $/yr 

37 Price of fish $4.30/kg (1.95/lb) 


38 Money supply in market for fish $187.5x10 



40 Tidal energies 0.1x10 FFE kcal/yr 

41 Fish stocks outside the bay ** 89.36x10 kg/yr 


42 Solar and wind energy 8.88x10 FFE kcal/yr 

43 Source of rain to hammocks and pinelands 98.2 inches/yr 

190,620 ac ft/yr 

44 Source of rain to slough marshes & wet prairies 98.2 inches/yr 

2,915,800 ac ft/yr 

45 Source of rain to mangroves & s.» , marshes 98.2 inches/yr 

3,852,770 ac ft/yr 

50 Money from state, local and flood control dist. 

due to park $6,400/yr 

51 Federal allocation to park 2,455,000 $/yr 

52 Visitors 1,017,393 visitors/yr 

53 Money spent in park attracted dev. by tourist -'*'\5. 5x10 $/yr 

54 Retail value of fish $23. 48xl0 & /yr 

55 Fish and shellfish yield 2.85x10 kg/yr 

56 Fuels by FCD on park $1 . 90/y r ; 13. 6x10 kcal/yr 

57 Electricity used by FCD $120/yr ; 2 . 06x10 kcal/yr 

58 Goods & services used by FCD $6090/y r ;64 . 8x10 kcal/yr 

59 Dollars for FCD for park $6,400 


60 Fuels used in park 3.9.4*10 kcal/yr 


61 Electricity to park 3.xl0 kcal/yr 


62 Goods & services to park from external economy 16x10 kcal/yr 

63 Fuels to fishing $1,135x10 /yr or 

79.45xl0 9 kcal/yr 

64 Goods & services to boats $12. 35xl0 & /yr or 

185. llxl0 9 kcal/yr 

65 Money spent on boat operations $12.26x10 /yr 


66 Fuels used by park attracted development ** 46gxl0 ffe kcal/yr 


67 Electricity used by park attracted development ** 284x10 ffe kcal/yr 


68 Goods & service- to park attracted dov°lopment ** 750x10 ffe kcal/yr 

69 Money spent by park attracted development ** 45.5x10 $/yr 


70 Image and information from parks productivity and 8070x10 ffc/yr 
diversity 412xl0&$/yr 

71 Mosquito impact on image ** -»351. 2x10 $/yr 


Footnote it 

7 2 




Contributions of park attracted dev. to park image 

Contribution of image to attract visitors 

Income from park fees, etc. 

Cost of goods & services, Fuels, etc. 820,000 $/yr 

Park money to park attracted development 2,071,000 $/yr 

(alculations and sources in Appendix Table A- 1 , 

Values based on particularly sparse data. 

FIGURE 4b. Natural systems, Everglades National Park. 







FIGURE 5. Details of Everglades National Park related dollar and energy flows, 1975, 



a. Everglades National 
Park only. 




Energy investment ratio = jj = 0.008 



Energy investment ratio = jj = 0.28 

b. Everglades National Park, park attracted development and related fishing 
industry . 

x 10 9 CE kcal/yr 

P = Purchased energies for external economy 

N = Natural energy (solar energy) incident 
on park. 

FIGURE 6. Investment ratio of Everglades National Park and attracted development including 
related fishing industry. Based on land use data 1973, economic data 1974-1975 years. Natural 
energy was estimated as 4.56 X 10" CE ac/yr X 1,397,000 acres; purchased energies in a. were 
estimated as sum of values for input energies #'s 60, 61, 62b. Fig. 1. Purchased energies 
in b. were sum of energy flow #'s 56, 57, 50, 61, 63, 64, 66, 67. in Fig. 1. 




(j[717om "-VV 

To all <i .!,*i 


I. Upper Kissimmee Lakes 
2 Kissimmee River 
3. Loke Okeechobee 
4 Eastern Sandy Flatlands 

5. Everglades Agricultural Area 

6. Water Conservation Areas 
7 Coastal Ridge 

8. Everglodes National Park 

9. Big Cypress Swamp 

10. Western Sandy Flatlands 


O Greater than 1000 acre ft/ yr 

; > 500-1000 acre ft /yr 

> 100-500 acreft/yr 

— 10-100 acre ft/yr 

FIGURE 7. Hydrologic diagrams of the Kissimmee Everglade regions (data from Gayle et al. 
in Odum et al., 1975; and Carter et al.). 

f Boundory 

\ /"""Canal 


/ \/ 


/ \ 

A. \ 

yV x - 

^ X \ 

\ ^X N 

) X \ 



X J— ^ 



^^ YJ 

V. / / 

^n/ / 

A / 

.— - — -" — ^A / * 

°-\ 1 


°A i * 

o W 

\ *>\ i 

3 U 

\ c\ 1 *" 

■£\ °A 

a r\ 

\ \ * 


_Example ^\ / 



Mile Wide "A J *■ *- 
Fioodway *e>\ , 




' M. 

ik * A A A A. A A 


- a . j SCALE 


FIGURE 8. Everglades agriculture area showing an 
example of mile wide fioodway to convey water from 
Lake Okeechobee to water conservation areas and 
Everglades National Park. 


The natural system energy inputs are 6.4x10 
CE kcal/yr of sunlight, 2.51xlOJ-2 C E kcal/yr of 
wind and 2.5xl0 12 CE kcal/yr of rain (as chemical 
potential to ocean salinity) totaling 11.4 trillion 
(CE kcal) annually. This amounts to 8.2% of the 
natural energy input to the Kissimmee-Okeechobee 
basin (Browder, Littlejohn, and Young, 1976) . 

In 1975 about 1,017,393 visitors came to 
Everglades National Park bringing an estimated $45 . 5 
million dollars to the area around the Park (flows 
52 and 53 in Fig. 2a) . This represents 1% of the 
total regional tourist dollar income, and about 
10% of the tourist dollars of Dade County (Miami- 
Metro. Dept. of Publicity & Tourism, 1976) . An 
additional 23.6 million dollars are generated as 
retail sales of fish and shellfish. Values of 
sport and commercial fish and shellfish include 
wholesale shrimp value at the Dry Tortugas shrimp 
grounds, which are dependent on the Park. How- 
ever, if the Park nurseries were not there, it is 
estimated, that 75% of the shrimp catch worth 
$37.6 million would be lost to the region's 
economy. (See footnote 54, Table A-l) . The na- 
tional park system contributed an additional 
$2.5 million in 1975 for park operations, mostly 
salaries for people working at the Park. Money 
provided by tourists, fishing, and the government 
contribute to the economy of the area surrounding 
the Park, providing funds for the purchase of goods 
and services for the urban system from outside the 

Until June 1975, vegetab 
tomatoes, were grown within 
The 19 7 3-74 farm crop value 
minion. une value of the 
cent agribusiness (packing 
estimated to be $20 million 
1975) . Closing the "hole i 
cultural activity resulted 
related economy. However, 
water quality, protection o 
"image" by the action have 

le crops, primarily 
the Park boundary, 
s were estimated at $7 . 7 
tomatoes to the adja- 
and marketing) was 

(Cornwell and Atkins, 
n the donut" to agri- 
in loss of 24% of Park 
the possible gain in 
f wildlife, and Park 
not yet been quantified. 

Inputs From National Fossil Fueled Economy 

Fuel consumption for the Park f 
ing industry and Park attracted de 
63, 66, Fig. 3) was estimated at 5 
CE kcal/yr. This plus 287 billion 
electricity totaled 840 billion CE 
energy consumed, representing 0.4 3 
energy consumption and about 1% of 
fuel consumption. The energetic v 
and services used in the Park (#62 
951 billion CE kcal/yr, represents 
regional use of goods and services 
The total energy of fuels and good 
used in the Park was estimated at 
CE kcal/yr. 

acility, fish- 
velopment (#60, 
34 billion 

CE kcal/yr of 

kcal/yr of 
% of regional 

Dade County 
alue of goods 
, 64, 68, Fig. 3), 

0.44% of the 

in energy terms, 
s and services 
1.78 trillion 

Energy Investment Ratio 

Flows of ene 
aggregated in F 
(purchased ener 
Park alone (Fig 
Park developmen 
Park alone was 
the Park influe 
equivalent of s 
natural systems 
or pumping 0.28 
generated goods 
This is much le 
whole . 

rgy in coal equivalents are further 
ig. 6. An energy investment ratio 
gy/energy) was calculated for the 

6a) and for the Park and related 
t (Fig. 6b) . The ratio for the 
0.008, while that for the Park and 
need economy was 0.28. Each coal 
olar energy captured by the Park's 
was associated with "attracting" 
coal equivalents or £< ssil fuel 
and services into the region, 
ss than the 2.5 for the U.S. as a 

The Value of Water as an Amplifier of the Flow of 
Energy in Everglades National Park 

In an energy system, each input can be 
considered to amplify the flow of another, i.e., 
act as a multiplier. It is common knowledge that 
water is a very sensitive factor in plant produc- 
tion, and this is particularly true for wetlands. 
A calculation was made for the value of water in 
plant production based on an empirical relation 
of net primary production in g/n\2-yr to precipita- 
tion in mm/yr (Leith and Whittaker, 1975). 

The energetic and economic value of 1 ac-ft of 
water in a region with an annual precipitation of 
45-60 inches was estimated to be 0.666x10 CE/ac- 
ft or $34/ac-ft. This estimate is close to the 
$30/ac-ft for human use used by the U.S. Army 
Corps of Engineers in benefit-cost analysis. Con- 
trol of the Park's surface water supply lies out- 
side the Park's boundaries. South Florida's 
water management system and related problems are 
described in the South Florida report by Browder 
et al. (1976) . 

1000 mm/yr = 1400 g/m 2 /yr 

1500 mm/yr = 1800 g/m /yr 


400 g/m /yr 

5 00 mm/yr 

1=0.8 g/m 

x 10 mm/cm = 8 g/m -cm 

x 4.5 kcal/gm 



m /ac 








r ,4 









in/ft = 




10 6 


36 kcal/m -cm npp 

x 3 gpp/npp = 108 kcal/m -cm gpp 

x 0.05 CE/square = 5.4 CE kcal/m -cm gpp 

CE kcal/ac-cm 

CE kcal/ac in gpp 

x 10 CE kcal/ac-ft 

r 19,600 kcal/$ = 34.00$/ac-ft value of water 
for gpp in the range of 
rainfall, for E.N. Park 

If the water supply to the park were increased 
by routing 500,000 additional acre feet of water 
south through the Everglades, then the amplifying 
effect of water could cause the productivity of 
the Park's ecosystems to increase by a value 
equivalent to $17 million/yr ($34/acre-ft x 
500,000 acre ft/yr) . This value should be 
realized in the increased availability of food 
for wildlife and in an increased ability of the 
park to attract visitors. It also should improve 
the fish production in central Florida Bay where 
present salt concentrations of 41 to 51 ppt are 
reflected in reduced standing crops of fish 
relative to other parts of the bay (T. Schmidt, 
x976) . 

The ratio of fuel, goods and services to park 
related development, to the Park's GPP is 0.221; 
therefore, by the investment ratio hypothesis, an 
increase of $17 million/yr in natural production 
could generate $3.76 million/yr additional ma- 
terial from the external economy (0 . 221xl7xl0 6 ) , 
an increase of 5.1%. 


alculation of water amplifier values in 

ry production (values from Leith and Whittaker 


Assessment of Model and Evaluations 

This first attempt to estimate the flows of 
energy and money is likely to be underestimated. 
Among the pathways of interaction that we know a- 
bout that are not yet estimated are: 

1. Energy involvements of citizens who 
concern themselves with the park while 
using their energy-valuable time else- 
where . 

2. Portions of the budgets of Federal, 
State, and Local institutions concerned 
with the Park beyond that spent directly 
and locally. 

3. The total value of the sport fishery 
to the economy. 

4. The value of the Park to sport and com- 
mercial fisheries of the Gulf as well as 
those of Florida Bay. Energy value of 
the Park as a source region affecting 
southwest wind to the metropolitan 
section of Miami in contributing to 
rainfall, air purity, and temperature. 

5. Role of the Park in blocking over- 
development and lowering energy 
effectiveness of suburban economy. 

6 . Role of the Park in providing food and 
habitat to migrating water fowl, wading 
birds, and Passerines whose perpetuation 
is important to recreational activities, 
forest insect control, and other as- 
pects of natural ecosystems throughout 
the eastern seaboard. 

Poss ible Meaning of Low Investment Ratio in 
Present Calculations 

A full discussion will await integration of 
these results with the rest of the South Florida 
Study in the forthcoming final version of the 
Technical Report. 

The low ratio of purchased to natural energies 
may be interpreted by more than one hypothesis. 
As suggested in the previous paragraph the economy 
involved in the Park may be much underestimated. 
If, however, the true involvement of the economy 
of purchased attracted energy flow is less than 
the average in Florida, then one might expect to 
see more kinds of development involving the Park 
locally and at a distance. 

It may be important to compare the investment 
ratio of the Everglades Park with those like 
Yellowstone that may be more heavily visited. The 
ultimate question is what is the carrying capacity 
of the Park in supporting economy. And vice versa, 
what is the economic involvement which will suf- 
ficiently impress the public decision process to 
maintain public support and backing for the parks 
in future periods when there may be declining 
available energy and real GNP? 

A Suggestion to Restore a Corridor of 
"River of Grass" 

The timing and quantity of water released to 
Everglades National Park from the Everglades basin 
is thought to have changed significantly since 
the drainage of the Upper Everglades (Everglades 
Agricultural District) and the diking of the 

Central Everglades (Everglades Conservation Areas 
1, 2, and 3). Wildlife depending on the aquatic 
system have decreased by an estimated 89 percent 
since 1947, suggesting that, on a long-term basis, 
the Park receives less water than it received 
historically (Kushlan and White, 1975), although 
350,000 ac-ft annually was guaranteed when the 
present water control system north of the Park 
was implemented. 

Presumably the decline in wildlife is due to 
fire death and stress in life cycles due to al- 
tered hydroperiods , such reduction in the primary 
productivity that is the base of many food chains. 

Parker (1974) estimates that 
volume of the lower Everglades b 
equal to the outflow of Kissimme 
Lake Okeechobee (roughly one mil 
7a) . The excess water from Lake 
mally drained by sheet flow over 
and through the Caloosahatchee R 
through the Caloosahatchee was i 
only when lake stages were high 
Everglades was constrained from 
Atlantic by the Atlantic Coastal 
vegetation-clogged rivers in the 
Thus, the predominant course of 
down the length of the Everglade 

the outflow 
asin was roughly 
e River into 
lion ac-ft, Fig. 

Okeechobee nor- 

the Everglades 
iver. Drainage 
ntermittent and 

Water in the 
draining into the 

Ridge and the 

transverse ridges, 
surface flow was 
s and into Florida 

Over the last 10 
reduced the flow of 
into Florida Bay by 
was accomplished by 
canals from levees 
directing the water 
reach the Everglade 
ted to regulate lak 
irrigate and drain 
to maintain water 1 
areas. On the aver 
water leaves annual 

years drainage works have 
water out of the Everglades 
1/5 to 1/2 (Fig. 7b) . This 
dredging numerous large 
from Lake Okeechobee and 

flow to sea before it could 
s. These works were construc- 
e stages in Okeechobee, to 
agricultural districts, and 
evels in the water conservation 
age over 1.5 million ac-ft of 
ly through canals. 

The three major canals 
lets leading from the lake 
servation Areas now are us 
runoff north to the lake ( 
impossible to release wate 
into the Everglades during 
The water instead must be 
the Caloosahatchee and St. 
tain "safe" levels in the 
complete discussion of thi 
et al. , 1976) . 

that are the only out- 

to the Everglades Con- 
ed to pump agricultural 
Fig. 7b) . This makes it 
r from the lake south 

most of the wet season, 
released to sea down 

Lucie canals to main- 
lake. (See a more 
s problem in Browder 

Water "wasted" to sea by present drainage 
practices could be redistributed to the Everglades 
drainage basin to enhance natural productivity 
over much of South Florida by forming a shallow 
Everglades "floodway" between Lake Okeechobee and 
Conservation Area 3. The floodway could be 
diked to provide protection from high water in 
urban and agricultural areas. Based on Parker's 
(1974) estimates of the natural water supply to 
Florida Bay, around 500,000 ac-ft more water than 
the present average could be rerouted through the 
glades by such an outlet (Fig. 7c) approximating 
the natural situation (Fig. 7a) . 

A broad shallow floodway from the lake to 
Conservation Area 3 through the agricultural dis- 
trict could almost eliminate the loss of water 
through the Caloosahatchee and St. Lucie except 
in extremely wet years, and would increase the 
flow of water to Everglades National Park. The 
floodway would partially restore the natural pat- 
tern of water flow through the Everglades. Proper 
design to encourage the growth of emergent 


vegetation might lead to the regeneration of rich 
organic soils to replace those which are being 
depleted through biochemical oxidation in the 
drained agricultural fields (Stephens, 1974). 
Users on either side could pump into and out of 
the swale. Agricultural nutrients would become 
a benefit rather than a liability when back- 
pumped into Lake Okeechobee. 

An alternation of floodway and farmland over 
vertical (north-south) sections of the upper 
Everglades with reversal of roles of different 
vertical sections every 50 to 100 years, might 
greatly extend the productive life of the agri- 
cultural lands while at the same time restoring 
the natural north-south sheet flow through the 
Everglades . 

Another possible design was proposed by the 
Army Corps of Engineers in 1955. They suggested 
a mile wide "floodway" from Lake Okeechobee to 
Conservation Area 3 (Alignments A or B, Plan 6 
Floodway, Army Corps of Engineers, 1955) , like 
the one pictured in Fig. 8. The Corps estimated 
that the mile-wide floodway would have the 
capacity to carry 16,800 ft 3 /sec at the lake and 
23,040 ft 3 /sec at Conservation Area 3. At maxi- 
mum flow for 100 days, the floodway could carry 
approximately 320,000 ac-ft of water from the 
lake to Conservation Area 3 at a lake stage of 
17.6 ft. MSL. Emergent vegetation in the flood- 
way would slow water flow considerably, and the 
floodway should be wider to carry this much 
water and serve its other roles. 


The energy analysis of the Everglades Park 
suggests ways of measuring contributions of 
parks to the economy and their relative inter- 
action. Comparing different parks may be useful, 
using models for comparative analysis. What is 
the range in investment ratios when heavily 
visited parks are included? Is the ratio a mea- 
sure of economic impact on parks? The evaluation 
of energy quality of inputs suggests their rela- 
tive importance. The very high energy cost and 
amplifier effect of fresh water in South Florida 
indicates why it was used on land in biological 
productivity and can be again if a small river 
of grass can be restored as a corridor serving 
agriculture, urban development, and the Park. 


Appendices containing details of calculation, 
sources, and relevant literature will be pub- 
lished in the final technical report for the 
South Florida Project, Phase II, Center for 
Wetlands, University of Florida; or they will be 
available from the U.S. Dept. of the Interior, 
National Park Service, Contract report PX-0001- 
60080, Nov. 1976. Washington, D.C. 


March 28, 1955. Partial Definite Project 
Reports for Flood Control and Other Purposes, 
Part IV - Lake Okeechobee and Outlets, Supple- 
ment 2 Hydrology and Hydrologic Design. Sec- 
tion 5A - Design memorandum additional lake 
regulation facilities. 

BAYLEY, S., and H. T. ODUM. 1976. Simulation of 
interrelations of the Everglades' marsh, peat, 
fire, water, and phosphorus. Ecological Model- 
ling 1:169-188. 

South Florida: Seeking a Balance of Man and 
Nature. Bureau of Comprehensive Planning, 
Division of State Planning, Florida Dept. of 
Admin.; and National Park Service, Department 
of the Interior. Center for Wetlands, Univ. 
of Florida, Gainesville. 

CORNWELL, G. W., and K. ATKINS. 1975. The im- 
pact of evicting farmers from Everglades Na- 
tional Park's hole-in-donut . Ecoimpact, Inc. 
Gainesville, Florida. 

KEMP, M. 1977. Energy Analysis and Models of 
Power Plants in Coastal Zone. Ph.D. Disserta- 
tion. Dept. of Environmental Engineering 
Sciences, Univ. of Florida, Gainesville. 

KUSHLAN, J. A., and D. A. WHITE. 1975. A survey 
of wading bird populations in southern Florida 
1974-1975. Unpublished resource management 

LEITH, H., and R. H. WHITTAKER. 1975. Primary 
Productivity of the Biosphere. Springer- 
Verlag, New York. 

ODUM, H. T. 1971. Environment, Power, and So- 
ciety. Wiley-Interscience, New York. 

ODUM, H. T. et al. 1976. Net energy analysis of 
alternatives for the United States. Pages 254- 
304 in: U. S. Energy Policy: Trends and 
Goals. Part 5 - Middle and Long Term Energy 
Policies and Alternatives. 94th Congress, 
2nd Session, Committee Print. Prepared for 
the Sub-committee on Energy and Power, of the 
Committee on Interstate and Foreign Commerce 
of the U.S. House of Representatives. 66-723. 
U. S. Govt. Printing Office, Washington, D.C. 

ODUM, H. T., and M. T. BROWN (eds.). 1975. 

Carrying Capacity for Man and Nature in South 
Florida. Center for Wetlands, Univ. of 
Florida, Gainesville. 

ODUM, H. T., and E. C. ODUM. 1976. Energy Basis 
for Man and Nature. McGraw-Hill, New York. 

SCHMIDT, T. W. 1976. Florida Bay fisheries 
ecology. Interim report EVER-N-36-64 . 

ZUCHETTO, J. 1975. Energy-economic theory and 
mathematical models for combining the systems 
of man and nature, case study: the urban 
region of Miami, Florida. Ecological Modelling 

4 i 


11 2 

John E. Randall , Helen A. Randall , and Alan H. Robinson 

The Virgin Is 
was established 
self has much na 
around it that i 
ing this, the Na 
marine biologica 
(now the Rosenst 
pheric Science) 
was supported by 
Service (Dingell 
Virgin Islands) , 
the National Par 

lands National Park on St. John 
in 1956. Though the island it- 
tural beauty, it is the sea 
s the most captivating. Recogniz- 
tional Park Service initiated a 
1 survey with the Marine Laboratory 
iel School of Marine and Atmos- 
of the University of Miami. This 
the U. S. Fish and Wildlife 
Johnson appropriation to the 
National Science Foundation, and 
k Service. 

The survey commenced in November 1958 and 
ended on June 30, 1961. The research was carried 
out at a combined residence-laboratory provided 
by the National Park Service at Lameshur Bay on 
the south shore of St. John. The Randalls were 
present during the entire survey period as was 
Gladston Matthias, an islander who assisted in 
fishing and other operations. The first year's 
diving assistant was Herman E. Kumpf , the second 
year's James R. Chess, and the third Robert E. 
Schroeder. Gilbert L. Voss, C. Richard Robins, 
and Lowell P. Thomas engaged in project activities 
at the University of Miami such as the curating 
of marine biological specimens. 

Among the research projects undertaken during 
the survey were the preparation of a chart of the 
marine environments of St. John to the 10-fathom 
curve (Kumpf and H. Randall, 1961) , a survey of 
commercial and sportfishing in the waters of the 
island (Idyll and J. Randall, 1959) , the tagging 
of reef fishes for the study of growth and move- 
ments (J. Randall, 1962; 1963b), a study of the 
food habits of West Indian reef fishes (J. Randall, 
1967; J. Randall and Hartman, 1968) , the effect of 
grazing fishes on marine plants (J. Randall, 1961; 
1965) , analysis of fish populations on natural and 
artificial reefs in Lameshur Bay (J. Randall, 
1963a) , observations on the spawning and develop- 
ment of parrot-fishes (Randall and Randall, 1963), 
a study of the biology of the queen conch (J. 
Randall, 1964), a comparable study of the West 
Indian topshell (H. Randall, 1964) , research on 
the biology of the sea urchin Diadema antillarum 
(J. Randall, Schroeder, and Starck, 1964), and ob- 
servations of the predation by mongooses on the 
eggs of sea birds and the green turtle (Seaman 
and J. Randall, 1962) . 

Many systematic papers on fishes which were 
based wholly or in part on specimens from St. John 
have been published by J. Randall and associates, 
particularly C. Richard Robins and James E. Bdhlke, 
culminating in the book Caribbean Reef Fishes 
(J. Randall, 1968). Systematic and ecological 
studies on marine invertebrates from St. John ma- 
terial have been completed by Lowell P. Thomas, 
Gilbert L. Voss, Raymond B. Manning, and Anthony 
J. Provenzano, Jr. 

Bernice P. Bishop Museum, Fish Division, P.O. 
Box 6037, Honolulu, HI 96818. 
National Park Service, Denver Service Center, 

P.O. Box 25287, Denver, CO 80225. 

Lameshur Bay, St. 
scientists- in- the- se 
and Tektite II in 19 
with the support and 
National Aeronautics 
Department of Health 
Coast Guard, the Nat 
Smithsonian Institut 
Virgin Islands, the 
the Tai Ping Foundat 
Research was the pri 
Tektite I and the De 
Tektite II. 

John, was chosen for the 
a programs -- Tektite I in 1969 
70. These were carried out 
cooperation of the U. S. Navy, 
and Space Administration, 
, Education, and Welfare, the 
ional Science Foundation, the 
ion, the Government of the 
General Electric Company, and 
ion. The Office of Naval 
ncipal supervising agency of 
partment of Interior for 

During Tektite I four aquanauts lived and con- 
ducted research underwater for 60 continuous days. 
This was the longest saturation dive that has been 
undertaken. Their base was a habitat of two cyl- 
inders 4 m in diameter and 6 m high joined by a 
flexible tunnel and seated on a rectangular base 
at a depth of 15 m. The men made excursions from 
the habitat up to 500 m or more, but remained with- 
in the depth range of 7 to 28 m. In addition to 
their own reseach, they served as subjects for 
comprehensive medical and behavioral studies. 

Tektite II was a seven-month program consisting 
of ten different missions. Each mission which 
lasted from 10 to 20 days, involved four scientists 
and one engineer. 

The Tektite programs were eminently successful, 
resulting in some fundamental marine biological 
research. Two Science Bulletins of the Natural 
History Museum, Los Angeles County, have been de- 
voted to the results of these programs (Bulletin 
14, 1972, edited by Bruce B. Collette and Sylvia 
A. Earle and Bulletin 20, 1975, edited by Sylvia 
A. Earle and Robert J. Lavenberg) . Individual re- 
search papers within these bulletins are (from 
Bulletin 14): Sylvia A. Earle - The influence of 
herbivores on the marine plants of Grant Lameshur 
Bay, with an annotated list of plants; Thomas J. 
Bright - Bio-acoustic studies on reef organisms; 
Conrad Mahnken - Observations on cleaner shrimps 
on the genus Periclimenes ; F. G. Hochberg, Jr. 
and Robert J. Ellis - Cymothoid isopods associ- 
ated with reef fishes (abstract) ; William L. High 
and Alan J. Beardsley - The behavior of reef fishes 
in relation to fish pots (abstract); H. Edward 
Clifton and Ralph E. Hunter - the sand tilefish, 
Malacanthus plumieri , and the distribution of 
coarse debris near West Indian coral reefs; Ann 
C. Hartline, Peter H. Hartline, Alina M. Szmant 
and Arthur 0. Flechsig - Escape response in a 
pomacentrid reef fish, Chromis cyaneus; Bruce B. 
Collette and Frank H. Talbot - Activity patterns 
of coral reef fishes with emphasis on nocturnal- 
diurnal changeover; C. Lavett Smith and James 
C. Tyler - Space resource sharing in a coral reef 
fish community; (from Bulletin 20) : David A. 
Olsen, William F. Herrnkind and Richard A. 
Cooper - Population dynamics, ecology and behavior 
of spiny lobsters, Panuli rus argus , of St. John, 
U.S. V.I. (I) Introduction and general population 
characteristics; David A. Olsen and Ian G. Koblic - 
Population dynamics, ecology and behavior of spiny 
lobsters, Panulirus argus, of St. John, U.S. V.I. 
(II) Growth and mortality; Richard A. Cooper, 
Robert Ellis and Steven Serfling - Population 
dynamics, ecology and behavior of spiny lobsters, 

■\ : , 

Panulirus argus , of St. John, U.S. V.I. (Ill) 
Population estimation and turnover; William 
F. Herrnkind, John A. VanDerwalker and Louis Barr - 
Population dynamics, ecology and behavior of 
spiny lobsters, Panuli rus argus, of St. John, U.S. 
V.I. (IV) Habitation, patterns of movement and 
general behavior; Louis Barr - Biology and be- 
havior of the arrow crab, Stenorhynchus seti- 
cornis (Herbst) , in Lameshur Bay, St. John, 
Virgin Islands; Charles Birkeland and Brian 
Gregory - Foraging behavior and rates of feeding 
of the gastropod Cyphoma gibbosum (Linnaeus) ; 
C. C. Lee, E. L. Lee and J. S. Bunt - Distribu- 
tion of biomass in a coral reef transect; Arthur 
C. Mathieson, Richard A. Fralick, Richard Burns 
and William Flahive - Phycological studies during 
Tektite II, at St. John, U.S. V.I. 

Since 1966, the College of the Virgin Islands 
has operated the Virgin Islands Ecological Re- 
search Station at Lameshur Bay, St. John. In 
addition to playing host to the two Tektite pro- 
jects, the station has provided support and fa- 
cilities for projects investigating ciguatera 
(fish poisoning) and the development of commercial 
fisheries operations in the waters around the 
Virgin Islands. Results of these studies and 
other progress reports are available in a series 
of Contributions which may be obtained from the 
station's Director via Cruz Bay, St. John, U.S. 
V.I. 00830. The station, following a temporary 
closure in 1975-76, is available to student groups 
as well as long-term individual researchers. 

Alan H. Robinson was in residence on St. John 
from 1970 through 1974 as the marine research 
biologist with the National Park Service. His 
efforts were divided between actual research, the 
identification of basic resource management prob- 
lems, and the planning of specific inventory and 
research projects (National Park Service, 1975a; 
1975b) . Principal publications and activities 
included analysis of visitor damage along St. 
John and Buck Island's two underwater nature 
trails (Robinson, 1973), a two-year comprehensive 
analysis of coral-sand beach dynamics (Hoffman, 
Robinson and Dolan, 1974), an investigation of the 
development mechanisms and ecologic functions of 
saltponds on St. John (Robinson and Feazel, 1974), 
and an overview of research and resource manage- 
ment problems in the Virgin Islands National Park 
during the period 1970 through 1974 (Robinson, 
1976a). The National Park Service, in cooperation 
with the College of the Virgin Islands' Ecological 
Research Station on St. John, provided major sup- 
port for a variety of successful dissertation 
candidates from mainland institutions. A recent 
2*5-year study resulted in the most comprehen- 
sive quantitative analysis to date of fringing 
reef development on St. John (Feazel, 1975). 

Several general publications and marine 
conservation initiatives, not strictly reporting 
research, but drawing substantially from combined 
research efforts from 1959 to 1975, have also 
appeared recently. These include a popular-style 
interpretive booklet "Virgin Islands National 
Park - The Story Behind the Scenery," with photog- 
graphy by Fritz Henle (Robinson, 1974) . A summary 
paper in marine park planning is largely based 
upon observation and study of visitor activities 
and impacts in the Virgin Island park units 
(Robinson, 1976b) . Substantial progress has been 
made, based also on surveys incidental to the 
research cited above, in establishing the first 
marine reserve within the park system of the 
Territory of the Virgin Islands - a coral reef 
reserve at Lagoon Point, St. John. On a broader 
scale, the research staff at the West Indies 
Laboratory of Fairleigh Dickinson University on 

St. Croix has recently completed an important 
natural area survey for the National Park Service's 
Natural Landmarks program (Adams, Gerhard, Ogden 
and Bowman, 1975). This natural region theme 
study has identified some 14 sites on the three 
U.S. Virgin Islands as potential natural land- 
marks, including terrestrial coastal valleys and 
shoreline, offshore bird islands, and classic 
coral-bounded bays. Although eventual official 
designation as landmarks would not automatically 
provide for public ownership or unqualified pro- 
tection, the attention focused on these critical 
areas will go far to stimulate local government 
as well as private conservation/preservation 

The West Indies Laboratory of Fairleigh 
Dickinson University is currently undertaking a 
study of natural and man-induced changes in the 
ecosystem of Buck Island Reef National Monument 
with the support of the National Park Service. 
The principal investigators are W. B. and E. H. 
Gladfelter, R. K. Monahan, J. C. Ogden, and 
R. F. Dill. Buck Island is visited by approxi- 
mately 50,000 persons each year, many of whom 
snorkel on the underwater trail of the island's 
barrier reef. The Laboratory scientists will 
inventory the marine resources and determine the 
rate of accrual and the rate of destruction and 
modification of these resources. This will in- 
volve mapping the area, a census of fish popula- 
tions, study of the zonation of invertebrate and 
plant life, and the impact of human visitors. 

Since the early 1970s, a Bureau of Fish and 
Wildlife has been operating with the Territorial 
Government's Department of Conservation and 
Cultural Affairs. The Bureau's prime marine re- 
sponsibilities, under the leadership of long-time 
resident biologist Arthur E. Dammann, emphasize 
the concern to monitor and potentially improve 
the fisheries harvest of the Virgin Islands. 
This has included survey and analysis of extremely 
critical nursery grounds in the islands' remaining 
mangrove lagoons, artificial reef placement, and 
location of previously unexploited fish stocks at 
the edge of the islands' broad submarine shelf. 

Although much progress has been made in recent 
years in our understanding of the classification, 
natural history and inter-relationships of the 
marine fauna and flora of the Virgin Islands and 
elsewhere in the Caribbean Sea, much remains to 
be learned. The Virgin Islands National Park and 
Buck Island Reef National Monument off St. Croix 
are the only marine preserves of the United States 
in the Caribbean Sea. Because of the protection 
provided for marine life at these islands, a 
unique opportunity exists for research. It is, 
for example, far easier to make observations or 
take underwater photographs of fishes in areas 
where man's usual predatory activities are cur- 
tailed. It is hoped that the National Park 
Service will continue to encourage and sponsor 
marine biological research in the Virgin Islands. 



ADAMS, J. B., L. C. GERHARD, J. C. OGDEN, and 
J. BOWMAN. 1975. Potential national natural 
landmarks U.S. Virgin Islands. Virgin Islands 
Natural Region Theme Study prepared for the 
National Park Service by West Indies Labora- 
tory, St. Croix. 82 pp. 

COLLETTE, B. B., and S. A. EARLE . 1972. Results 
of the Tektite Program: Ecology of coral 
reef fishes. Natural History Museum of Los 
Angeles County, Science Bulletin 14:1-180, 

EARLE, S. A., and R. J. LAVENBERG . 1975. Results 
of the Tektite Program: Coral reef inverte- 
brates and plants. Natural History Museum of 
Los Angeles County, Science Bulletin 20:1-103. 

FEAZEL, C. C. 1975. Ecozonation and sediment 
distribution of three reef areas on St. John, 
Virgin Islands. Ph.D. Thesis, Johns Hopkins 
University, Baltimore, Md . 224 pp. 

HOFFMAN, S., A. H. ROBINSON, and R. DOLAN. 1974. 
Virgin Islands Beach Processes Investigation, 
St. John, Virgin Islands. Natl. Park Serv. , 
Office of Natural Science Studies, Occ. Pap. 
No. l:vii + 1-74. 

IDYLL, C. P., and J. E. RANDALL. 1959. Sport 

and commercial fisheries potential of St. John, 
Virgin Islands. Fourth Internatl . Gamefish 
Conf. (1959), 9 pp. + addendum (1960), 2 pp. 

KUMPF, H. E., and H. A. RANDALL. 1961. Charting 
the marine environments of St. John, U.S. Vir- 
gin Islands. Bull. Mar. Sci. Gulf and Carib. 
11(4) :543-551, 4 figs. 

NATIONAL PARK SERVICE. 1975a. Resource Manage- 
ment Plan for Virgin Islands National Park. 
Virgin Islands Natl. Park, St. Thomas, U.S. 
V.I. 101 pp., with projects in appendix. 

NATIONAL PARK SERVICE. 1975b. Resource Manage- 
ment Plan for Buck Island Reef National Monu- 
ment. Virgin Islands Natl. Park, St. Thomas, 
U.S. V.I. 36 pp., with project status sheet. 

RANDALL, H. A. 1964. A study of the growth and 
other aspects of the biology of the West Indian 
topshell, cittarium pica (Linnaeus). Bull. Mar. 
Sci. Gulf and Carib. 14 (3 ): 4 24-443 , 10 figs. 

RANDALL, J. E. 1961. Overgrazing of algae by 
herbivorous marine fishes. Ecology 42(4): 812. 

RANDALL, J. E. 1962. Tagging reef fishes in the 
Virgin Islands. Proc. Gulf and Carib. Fish. 
Inst. (14th Ann. Session, 1961 ): 201-241 , 8 
figs . 

RANDALL, J. E. 1963a. An analysis of the fish 
populations of artificial and natural reefs in 
the Virgin Islands. Carib. Jour. Sci. 3(1): 

RANDALL, J. E. 1963b. Additional recoveries of 
tagged reef fishes from the Virgin Islands. 
Proc. Gulf and Carib. Fish. Inst. (15th Ann. 
Session, 1962) : 155-157 . 

RANDALL, J. E. 1964. Contributions to the biol- 
ogy of the queen conch, strombus gigas. Bull. 
Mar. Sci. Gulf and Carib. 14 (2) : 246-295, 13 
figs . 
RANDALL, J. E. 1965. Grazing effect on sea 

grasses by herbivorous reef fishes in the West 
Indies. Ecology 46 (3) : 255-260 , 4 figs. 
RANDALL, J. E. 1967. Food habits of reef fishes 
of the West Indies. Stud. Trop. Oceanog. 
Miami 5:665-847. 
RANDALL, J. E. 1968. Caribbean Reef Fishes. 
305 pp., 319 illustr., TFH Publication, New 
RANDALL, J. E., and W. D. HARTMAN. 19 68. Sponge- 
feeiding fishes of the West Indies. Marine 
Biology 1:216-225. 
RANDALL, J. E., and H. A. RANDALL. 1963. The 
spawning and early development of the Atlantic 
parrot fish, Sparisoma rubripinne , with notes 
on other scarid and labrid fishes. Zoologica 
48(2):49-59, 2 pis., 2 text figs. 
II. 1964. Notes on the biology of the echi- 
noid Diadema antillarum. Carib. Jour. Sci. 
4(2 S, 3) :421-433, 3 figs. 
ROBINSON, A. H. 1973. Natural versus visitor 
related damage to coral reefs in Virgin 
Islands National Park and Buck Island Reef 
National Monument, with recommendations on 
underwater nature trails. Private printing. 
24 pp., illustrated. 
ROBINSON, A. H. 1974. Virgin Island National 
Park--The Story Behind the Scenery. 4 8 pp., 
with 102 color photos by Fritz Henle and 
Clarendon Bowman. K. C. Publications, Las 
Vegas, Nevada. 
ROBINSON, A. H. 1976a. Marine research and re- 
source management in Virgin Islands National 
Park. Trans. Natl. Park Centennial Symposium: 
Research in the parks, at AAAS meeting 28-29 
December, 1971. USDOI/NPS Symposium Ser. No. 
ROBINSON, A. H. 1976b. Recreation, interpreta- 
tion and environmental education in marine 
parks. Proc, I. U.C.N. Internatl. Conf. Mar. 
Parks and Reserves, working pap. no. 4:99-121. 
I. U.C.N. Publ. New Series No. 37. 
ROBINSON, A. H., and C. C. FEAZEL. 1974. 

Fringing reefs, enclosing bays and saltponds 
in St. John, Virgin Islands. Oceans 7(5):40- 
SEAMAN, G. A., and J. E. RANDALL. 1962. The 
mongoose as a predator in the Virgin Islands. 
Jour. Mammalogy 43 (4 ): 544-546 . 



Richard A. Dirks 

The climatology of a region is a critical com- 
ponent in the effective management of the natural 
environment for its utilization in recreation and 
tourism. Recognizing the importance of baseline 
climatic data to management and research, clima to- 
logical studies of Yellowstone and Grand Teton 
National Parks were initiated in 1973. Our re- 
search in the Parks has been conducted in two 
parts: (1) the assembly, summary, and interpre- 
tation of existing data in the Parks region, and 
(2) the collection, reduction, and interpretation 
of data from a special network of meteorological 
stations. Overall, this study is attempting to 
produce a comprehensive description of the climate 
of the Parks based on an evaluation of all avail- 
able climatic records. The only previous attempt 
at a climatic summary was made for Yellowstone 
Park by Fletcher in 1927. In addition, individ- 
ual station summaries have been published for 
several of the regular long term stations . 

Fragmentary records of meteorological measure- 
ments date back to some of the early exploration 
chronicles and Park Superintendent reports at 
Yellowstone Park. Fairly complete records were 
kept at the Park Headquarters at Mammoth during 
the period 1879-1881. Superintendent Norris sum- 
marized his impressions of the Yellowstone Park 
climate in his 1879 and 1880 annual reports, in 
which he represented the climate as somewhat of a 
local anomaly resulting from the heat and mois- 
ture emitted by the hot springs and geysers. 
Continuous climatological records began at Mam- 
moth in 1887 and represent the second longest con- 
tinuous record of climatic data in the State of 
Wyoming . Continuous records in the Grand Teton 
Park date back to 1911 at Moran. 

The region comprising Yellowstone National Park 
and Grand Teton National Park lies in the north- 
west corner of Wyoming with portions of the north- 
ern and western boundaries of Yellowstone Park 
extending into Montana and Idaho. Yellowstone 
Park is primarily a heavily forested volcanic 
plateau with an average elevation around 8,000 
feet. The plateau is generally surrounded by high 
mountains except for the Snake River High Plains 
on the southwest which are about 1,000 feet lower 
in elevation than the volcanic plateau. The Con- 
tinental Divide traverses the Park from the south- 
east corner in a west-northwesterly route so that 
approximately the southwest quarter of the Park 
represents headwaters of the Snake River drainage. 
Drainage to the north and east flows into the 
Yellowstone, Madison, and Gallatin Rivers. Some 
of the larger rivers, particularly the Yellowstone, 
have cut deep canyons into the lava plateau, in 
some places producing spectacular high waterfalls. 
At their lower reaches, both the Yellowstone River 
and the Madison River flow through re\atively 
broad valleys at elevations oelow 7,500 feet, 
dropping below 5,200 feet near Gardiner, the low- 
est point in the Park. The highest elevation in 
the Park occurs on Electric Peak (11,155 feet), 
about six miles west of Gardiner. 

The physiography of 
is characterized by the 
Mountains on the wester 
peaks exceeding 12,000 
range extends south-sou 
stone Plateau and varie 
miles. The eastern slo 
abruptly down to Jackso 
montane basin about 48 
miles wide . The Snake 
basin floor which varie 
to 7,000 feet. To the 
tional Park lie the Abs 
southeast lies the Gros 

Grand Teton National Park 

steep slopes of the Teton 
n boundary with several 
feet in elevation. This 
thwestward from the Yellow- 
s in width from 10 to 15 
pes of the Teton Range fall 
n Hole, a narrow inter- 
miles long and 6 to 12 
River meanders through the 
s in altitude from 6,000 
east of Grand Teton Na- 
aroka Mountains and to the 
Ventre Range . 

Due to the complexity of the region's mountain- 
valley topography and the frequent location of 
climate stations in locales of moderate climatic 
conditions (such as mountain valleys) , a complete 
description of the Parks' diverse climates is not 
possible. In particular, the mountainous terrain 
of the Parks is not adequately represented by 
existing data . 

Climate Description 

The following discussion of climatic elements 
is not intended as a complete description of the 
climate but rather to give an indication of its 
various features and, more importantly, to give 
examples of the types of climatic information 
which can be provided for applied needs. 

In general, the climate of the Parks is charac- 
terized by short, cool summers, usually pleasant 
open fall, and long, rigorous winters, which tend 
to linger late into spring. The severity of win- 
ter is the result of persistent, rather low tem- 
peratures and much cloudy, unsettled weather with 
frequent light snows. 

Tempera tare 

Tempera tur 
cal of high a 
gions . Since 
high elevatio 
the air is re 
temperature r 
are commonly 
40°F during s 
occur during 
Park region 

e conditions i 
ltitude inter i 
the entire Pa 
ns leeward of 
latively dry, 
ange . Daytime 
replaced by ni 
ummer. Subfre 
all summer mon 
except at elev 

n the Parks are typi- 
or mountainous re- 
rks region lies at 
major mountain ranges, 
with a large diurnal 

maxima of 70-80°F 
ghttime minima below 
ezing temperatures 
ths throughout the 
ations below 6,000 

Department of Atmospheric Science, University 
of Wyoming, Laramie. 

Daytime maximum temperatures during summer 
are closely aligned to elevation as seen in 
Figure 1. Below 7,000 feet July maxima often 
reach 80°F, but temperatures in excess of 90°F 
are rare. Daytime maxima rarely reach 80°F 
at high plateau and mountain stations (el. 
>7,500 feet) . 

Winters are cold with daily maxima often re- 
maining below freezing. January is the coldest 
month with temperatures typically ranging from 
near zero at night to the middle twenties in the 
early afternoon at nonmountain locations. 


Midwinter extremes can be very cold. The record 
low is -66°F at Riverside RS (West Yellowstone) 
on February 9, 1933, followed by -63°F on the same 
date at Moran. All areas of the Parks except 
Gardiner have recorded at least 40°F below zero. 
The occasional severe cold waves usually do not 
continue for more than two or three days; however, 
in December 1924, the minimum temperature at 
Mammoth Hot Springs did not rise above 26° below 
zero for five consecutive days and above 12° below 
zero for ten days . 

The cold winter conditions are primarily of lo- 
cal origin produced by the combined effects of 
nocturnal radiation from a persistent winter snow- 
pack through a clear, dry atmosphere. The drain- 
age of radiationally cooled air into the basins 
and valleys may result in quite cold temperatures 
continuing for several days until this air is 
mixed out by strong winds or winter storm condi- 
tions. Extreme cold conditions occur when Arctic 
air invades the sheltered intermountain basins and 
valleys, stagnates there, and is further cooled by 
radiation and slope drainage. Since gravitational 
effects cause the cold air to drain into lower 
elevation valleys and basins, and cold air masses 
are often quite shallow, temperatures on the high 
plateaus and lower mountain slopes are often con- 
siderably warmer than temperatures at lower ele- 
vations during extreme cold situations. Figure 2 
illustrates this situation from January daily mini- 
mum temperatures for three stations located along 
a valley slope near Gardiner: Stevens Creek at 
the valley floor (el. 5,450 feet), Hayes Ranch 
(el. 6,160 feet), and Parker Point (el. 7,830 
feet) . Under neutral lapse conditions, the tem- 
perature at Hayes Ranch should be about 4°F colder 
and at Parker Point 13 °F colder than at the valley 
floor. Note the occasional deep isothermal or 
inversion conditions. Similar differences in mean 
climatic conditions are evident between Gardiner 
(el. 5,300 feet) and Yellowstone Park Headquarters 
(el. 6,341 feet), which lies well up the valley 
slope. Valley stations at Lamar RS (el. 6,470 
feet) and Tower (el. 6,266 feet) are at about the 
same altitude as Yellowstone Park Headquarters 
but winter minimum temperatures are on the average 
8-10°F colder. 

West Yellowstone Basin has the dubious honor of 
recording the record minimum temperature for the 
Parks region, -66 °F on February 9, 1933, at River- 
side RS located just east of West Yellowstone. 
Average minimum temperatures during the winter are 
also among the lowest in the Parks region. 
Several factors seem to account for the occurrence 
of extremely low temperatures in this area. Cold 
air outbreaks which enter the valley are often 
stagnated because they must drain to the north. 
Secondly, being surrounded by high mountains on 
all sides, there is extensive down-slope drainage 
of cold air in the basin. Finally, the drainage 
flow out of the basin must pass through a narrow 
constriction near Hebgen Dam, which further re- 
tards cold air outflow. A similar situation ex- 
ists in Jackson Hole and accounts for extreme co\d 
conditions in that basin. 

Lake Yellowstone temperature data should be 
quite typical of the plateau region if adjusted 
for minor lake influences during summer and fall, 
and if local adjustments are made for elevations . 


The normal annual precipitation at standard 
climatic stations in the Parks ranges from about 
11 inches at Gardiner to 38 inches at Bechler 
River. Short term records suggest average values 
in excess of 50 inches over the southern 

Yellowstone Plateau and in the Snake River High 
Plains, the only area in the Park which does not 
lie in the rain shadow of a major mountain range. 
Even higher annual amounts undoubtedly occur in 
mountainous regions, since winter precipitation 
in particular is highly dependent upon elevation. 
Unfortunately, precipitation records do not exist 
for these areas . 

Recently, Fames (1971) has shown that mean an- 
nual precipitation for mountainous areas of Mon- 
tana could be closely estimated from April 1 snow 
water equivalent measurements if corrections were 
applied for the canopy cover at each snow course. 
By plotting precipitation as a function of ele- 
vation for various terrain profiles and exposures, 
isohyetal maps could be constructed. Figure 3 
shows the mean annual precipitation for Yellow- 
stone National Park as determined by Fames. Re- 
cent studies in Yellowstone National Park by 
Despain (1973) show that vegetation zones are well 
differentiated by mean annual precipitation and, 
furthermore, these zones are closely aligned to 
the precipitation analysis given by Fames. Pre- 
cipitation shows a large gradient from north to 
south in Jackson Hole. High values at Snake 
River exceed 30 inches annually, decreasing to 15 
inches at Jackson. 

The areas of high mean annual precipitation in 
the southern Yellowstone Plateau, the Snake River 
High Plains and mountains all show major increases 
in precipitation during mid-winter months. The 
remainder of the Parks region has a fairly uniform 
distribution of precipitation throughout the year 
with small peaks during June and mid-winter. 

A wide range in annual snowfall occurs in the 
Parks due to the combination of elevation and 
"rain shadow" effects. Lowest values are esti- 
mated near 50 inches of Gardiner and 80-90 inches 
at Yellowstone Park HQ . Annual amounts increase 
southward across the Yellowstone Plateau to 150 
inches at Lake Yellowstone, and in excess of 400 
inches over the southwestern portion of Yellow- 
stone National Park. The annual totals decrease 
to about 180 inches in the northern regions of 
Jackson Hole with further reduction to less than 
100 inches at Jackson. At higher elevations an- 
nual snowfall is often in excess of 400 inches 
and locally may reach 600 inches. A recent 
analysis of average annual snowfall (1958-1972) 
by Fames (1974) is reproduced in Figure 4. 

Except for the 
stone National Park 
heavy snowstorms ar 
protect most areas 
slope precipitation 
these protected are 
Yellowstone during 
Moose during Januar 
Yellowstone Park HQ 
uncommon, significa 
been measured durin 
the Parks except ne 

southwestern portion of Yellow- 

and the higher mountains, 
e rare since upwind mountains 
from immediate orographic up- 

Record daily snowfalls in 
as include 24 inches at West 
January 1962, 21 inches at 
y 1962, and 14 inches at 

on 24 October 1919. Although 
nt amounts of snowfall have 
g the summer months throughout 
ar Gardiner and in Jackson Hole. 

Because of the persistent cold temperatures 
during winter throughout most of the Parks re- 
gion, considerable accumulation of snow may oc- 
cur. Lowest amounts occur in the Gardiner- 
Yellowstone Park HQ region where there is little 
accumulation until mid-winter or later. Average 
depths in this region are around 12 inches at 
the end of February with a record depth of 35.6 
inches in March 1917 at Yellowstone Park HQ. 
Snow depths of 3-4 feet are not uncommon at West 
Yellowstone, Old Faithful and Moran, while at 
Jackson accumulations of 2-3 feet occur during 


8000 r 






o 6000 



70 BO 


FIGURE 1. Mean maximum July temperatures ( F) for 
Parks region stations plotted against station 
elevation (feet) . Solid line is dry adiabatic lapse 
rate (-5.4°F per 1,000 feet). 

Sr«Tl:ns 1 - STEVE.N S CKECK 

i — i — i — r — i — i — : — : — r— 

rEKK 137E MGNtn 1 

+ = n»rES fWNcn y • rwxt' POINT 

' i -i — ! — i — i — i — i — i — I — I — r ~i — i — i — I — i — i — r 


© +■ Y Y 

c v 


Y * 

+ Y J 

* + 

1- Y 


I I I l_ 

i 1. 1 I 

J L 

J l_ 

J. I 1 L 

J I I I L 

Z 3 4 S E 7 9 S II U II I! 14 IS li II II I] a 21 C S M S S !1 a ^3 JO 31 

DAY or Tnt ionth 

FIGURE 2. Daily minimum temperatires (°F) at the 
3-station transect near Gardiner, Montana. 




FIGURE 3. Mean annual precipitation (inches) for Yellowstone National Park 
(1953-67), by Fames (1973). 

5 2 

FIGURE 4. Mean annual snowfall (inches) for Yellowstone National Park (1958-72), 
by Fames (1971) . 

r ;i 

FIGURE 5. Seven-year seasonal and annual weighted mean precipitation 
at Lake Yellowstone (1904-70). 

uw -w i Nuri 


nuns i97i - n?6 mrnrn . i - j naKS , 2 . 

STATION NO. » 12 STATION finne > nflTtS KAW" 

ITtlWS 19'« - 197S MONTHS - 6 - 9 HOURS I 

rtuctNT cum = on z 


siatiun i«j- ■ is smnoN »me - urrtR stNOtrvouc 
tuw ia7S hunims - i - ? iolks le - i? 
rtm-tNT . uo z 

' V. s 


Figure 6 

1 1 

1 1 1 




>5 / 


>IO // 




1 l~~-~Oi- 

>IS /\ " 


Figure 7 

FIGURES 6-8. Frequency of occurrence of winds greater than or equal to the designated 
value (MPH) for each direction. 

r 4 

middle and late winter. High mountain snow courses 
frequently record depths in excess of 100 inches, 
and over 200 inches has been reported. 

In an effort to determine long-term climatic 
trends in the Parks region, several analyses of the 
precipitation data were carried out. Data from 
several long-term station records were analyzed 
using a weighted running mean. The seasonal 
breakdown of data used was given by: Summer - 
July, August; Fall - September, October; Winter - 
November, December, January, February, March; 
Spring - April, May, June. 

Lake Yellowstone annual precipitation (Figure 
5) reached a maximum around 1908, then decreased 
gradually to its lowest point in 1930 remaining 
there until 1935; it again began increasing 
gradually until 1943, after which time the precip- 
itation level has remained relatively constant 
except for a slight dip in the 1950's. Though 
winter precipitation contributes most of the total 
annual precipitation, long-term fluctuations in 
precipitation are reflected in all four seasons 
while short-term fluctuations generally are a 
function of changes in winter precipitation. 

Further comparisons were made with frequency 
of drought occurrence -- that is, periods of no 
precipitation >_ 15 days -- and the frequency of 
heavy precipitation events — _ 0.5 inches per 
day. In general, all of these indicators showed 
a correlation in Parks precipitation with the 
1930' s drought. Other drought periods reported 
in the Great Plains and elsewhere are not generally 
reflected in the Parks data; and in some cases, 
well-defined local drought conditions and heavy 
precipitation conditions are evident only at 
individual stations. 

No long-term trends are evident in the data 
from Yellowstone Park; however, there is some 
evidence of an increase in annual precipitation 
for Jackson Hole since the early 1950 's from data 
at Moran and Jackson. This trend is due to in- 
creases in winter and spring precipitation. 


Wind is undoubtedly the most difficult of the 
standard meteorological parameters to measure and, 
furthermore, it has the greatest spatial and tem- 
poral variability near the surface. Local effects 
such as topographic channeling, slope winds, 
valley winds, mountain waves and obstructions pro- 
duced by vegetation and terrain may frequently 
dominate the wind records at a particular site. 
Therefore, unless the exposure of a station is 
carefully examined, caution must be used in gen- 
eralizing from its records. 

Contemporary wind records in the Parks are 
sparse. Only fire weather stations, where obser- 
vations are limited to one per day during summer 
months, have a reasonably complete network and 
many of these stations have been discontinued in 
recent years. Hourly winds are currently reported 
only at Jackson and West Yellowstone in the Parks 

Reports of severe winds are infrecient. Wind 
speeds rarely exceed 40 mph at Yellowstone Park 
HQ . Maximum winds of 75 mph are reported for 
Jackson, occurring very infrequently. Moose, 
Moran, West Yellowstone and Lake Yellowstone have 
all reported occasional blow-down of timber due to 
damaging winds from intense thunderstorms. Strong 
winds from thunderstorms (60-80 mph) have also 
been reported occasionally at the fire lookouts un 

Mt. Washburn and Mt. Holmes during summer months. 
Regions of strong downslope chinook winds are not 
defined in the Parks although some of the late 
fall and springtime timber blow-down may be at- 
tributable to such phenomena. During winter per- 
sistent strong winds are common on upper mountain 
slopes . 

In order to better define the wind climate of 
the Yellowstone-Grand Teton Parks region, field 
studies of climatic wind conditions were initiated 
during the summer of 1974. Three recording wind 
and temperature instruments were installed along 
a slope transect near Gardiner, Montana, on a 
southwest facing slope above the Yellowstone River, 
as described earlier. A second slope transect is 
situated at Teton Village, Wyoming, on the south- 
east facing slope of Rendezvous Peak in the Teton 
Range. Telemetered wind and temperature instru- 
ments are operated midway (el. 8,500 ft.) and near 
the top of the peak (el. 10,300 ft.). A record- 
ing wind instrument is also located on an exposed 
area of the central Yellowstone Plateau near Hay- 
den Valley. 

A preliminary analysis of data will be pre- 
sented for several of these stations to illustrate 
the wide range of wind conditions and some of the 
characteristic conditions associated with certain 
topographic features . 

During winter Hayden Valley is a rather harsh 
environment both day and night. Nocturnal wind 
conditions show a preference for southwesterly 
flow with speeds generally less than 10 mph, al- 
though, calm conditions occur only 15% of the 
time. Mid-winter days show a persistent strong 
northwesterly flow with the strongest winds from 
the west, sometimes having sustained velocities 
of 25 mph (Figure 6) . 

In a sheltered basin such as Jackson Hole, winds 
are generally light and calm conditions occur 
frequently during all seasons. 

In contrast to Jackson, the wind data collected 
in the Yellowstone Valley near Gardiner show the 
classical characteristics of the mountain valley 
circulation. At night during summers (Figure 7) 
and both night and day in winter a weak down- 
stream flow of cold air is commonly observed. 
During summer days a well-defined valley breeze 
is evident. Only occasionally is the local cir- 
culation over-powered by passing storm systems. 

Characteristics of windflow in high mountain 
regions can be exemplified by the data at upper 
Rendezvous. During summer there is a dominant 
westerly and northwesterly flow both day and 
night, strong winds occur occasionally with 
average speeds 15 mph. During winter the wind 
speed increases somewhat with average speeds in 
excess of 20 mph and sustained speeds of 40-50 
mph are observed (Figure 8) . During winter the 
flow is generally northwesterly with some west- 
erly and southwesterly components and a weak 
secondary peak from the east. 


There are many areas of environmental manage- 
ment in which climatological results can be ap- 
plied to practical problems . A few examples 
include : 

1. Environmental influences of weather on 
plant and animal life. 

2. Research on the bioclimatology of in- 
sects and pests. 

5 5 

3 . Research on the bioclimatology of plant 
diseases . 

4. Meteorological influences on fire control, 
including controlled burning operations. 

5. Location and design of structures, build- 
ings and transportation routes. 

6. Selection and layout of recreational areas- 
campsites, winter recreation sites, etc. 

7. Water resource management for national 
park usage. 

8. Meteorological aspects of air pollution 
related on tourism and nearby external 
sources such as lumber industry, mineral 
exploration and power plants. 

These and other applied studies can be fruit- 
fully pursued only if an adequate long-term base 
of climatological data is available on which local 
short term studies can be built. To provide and 
maintain such a data base requires a well-designed 
network of climatological stations located at 
carefully selected, representative sites, with 
regularly maintained and calibrated instruments. 
An accelerated effort is recommended in establish- 
ing such data backgrounds in large national parks. 
The pristine environment of remote areas of large 
national parks also provides an opportunity for 
collecting basic background data on many atmos- 
pheric parameters which are not subject to con- 
tamination by developing nearby sources. From 

such data long term natural trends can be more ef- 
fectively assessed and natural responses of bio- 
logical and geological processes to weather ef- 
fects can be more readily determined. Such data 
collection is critical to both national parks 
management and national environmental management 
in general. Closely coordinated research between 
Parks scientists and managers and nearby uni- 
versity scientists would promote important op- 
portunities and benefits for both groups. 

As people and use pressures increase on many of 
our national parks' lands, increasing problems 
emerge in trying to maintain the quality of the 
natural environment. Maintenance of the environ- 
mental quality of our national parks requires that 
park administrators and scientific researchers 
cooperate to optimize environmental management pro- 
cedures, both now and in the future. 


DESPAIN, D. G. 1973. Majo 
Yellowstone National Par 
mation Paper No. 19, Yel 
Park Service. 

FARNES, P. E. 1971. Mount 
hydrology from snow surv 
the Western Snow Confere 
April, 44-49. 

. 1973. Prelimi 

annual precipitation for 
Park. Unpublished, Soil 
Bozeman, Montana. 

1974. Prelimi 

annual snowfall for Yell 
Unpublished, Soil Conser 
Bozeman, Montana. 

r vegetation zones of 
k. Unpublished, Infor- 
lowstone Park, National 

ain precipitation and 
eys. Proceedings of 
nee, Billings, Montana, 

nary analysis of mean 
Yellowstone National 
Conservation Service, 

nary analysis of mean 
owstone National Park, 
vation Service, 



1 2 

Wilfred D. Logan and F. A. Calabrese 


I (Wilfred Logan) have been asked to present a 
brief, general history of the Service's concern 
with archeology. This history will have limita- 
tions in that it will be general. It will also be 
spotty, for I will use only a few areas and a few 
situations as examples. Some assertions I will 
make are based on intuition only, for, at this 
stage of my research into the history of our dis- 
cipline in the Service, I cannot document every- 
thing. This history will be biased by my own ex- 
periences for I have played a part in much of 
what has transpired since World War II; sometimes 
unwittingly, sometimes with full knowledge as an 
enthusiastic activist. I give this warning of 
bias, for other interpretations of events are pos- 
sible. On the whole I will try to review here 
how we got started, how we operated, and how effec- 
tive we have been. 

Federal concern with archeology pre-dates the 
National Park Service by many, many decades. The 
Smithsonian Institution has sponsored studies of 
antiquities since the 1840's. Other early studies 
and observations were made by a variety of indivi- 
duals and groups — the Bureau of American Ethnology; 
both public and private institutions; private citi- 
zens, and concerned groups of interested laymen — 
who became increasingly alarmed over the fate of 
antiquities in the public domain. Some of these 
groups, both public and private, were engaged in 
research and the mitigation of impacts upon sites 
being destroyed or damaged by construction activi- 
ties. Their concerns contributed to the stream 
of thought that brought about, ultimately, passage 
of the Antiquities Act of 1906. For those inter- 
ested in details of this movement, Ronald F. Lee's 
(1970) presentation ably sets forth events leading 
to passage of this significant item of preserva- 
tion legislation. The Antiquities Act is not only 
important for preservation of prehistoric sites 
but is also important because it adumbrated the 
National Parks Act of 1916. The creation of 
National Monuments under the terms of the 1906 
Act, along with the existence of several National 
Parks prior to 1916, formed effective pre-conditions 
for an agency concerned with administration of 
lands set aside for educational and inspirational 
purposes . 

Passage of the Antiquities Act resulted from 
alarm over the unscientific exploitation of ruins, 
primarily in the West. In effect, the Act prohib- 
ited "mining" of sites on the public domain by non- 
scientists. The Service, then, when it came into 
being, fell heir to a variety of areas preserving 
prehistoric remains. It is surprising in some 
respects, considering the number of prehistoric 
properties that came to the new Service, that pre- 
history has not had a more prominent position and 
a more profound role in Service admini: tration. 

As I view the Service's management of prehistory, 
both in care of sites and in research, I see sev- 
eral continuing threads of thought and action. 
Just as the Antiquities Act was brought into being 

National Park Service, Denver Service Center, 
Denver, Colorado. 
National Park Service, Midwest Archeological 

Center, Lincoln, Nebraska. 

by the efforts of scientists not necessarily em- 
ployed by the Federal government, so was most 
early archeological work carried on by scientists 
from institutions not directly connected with the 
management of the sites. One such early example 
is the work of Jesse Walter Fewkes at Mesa Verde 
(Fewkes, 1909 and 1911) . After the creation of 
the National Park Service, this pattern persisted, 
and the early policy was one in which research, by 
agencies and institutions other than the Service 
itself, was encouraged. This is largely true today. 
While such an approach may, in part, be desirable, 
I believe heavy reliance on this policy has helped 
develop an attitude of minimal concern, on the part 
of Service officials, for cultural resources which 
happen to exist on Service lands. 

It is easy to see why such an approach should 
flourish. First, there is precedent, as there was 
no National Park Service when the early studies 
were carried on. Second, the tasks of the infant 
National Park Service were staggering with regard 
to organization; development of facilities; devel- 
opment of access; and development of an adequate 
budget. It would be natural to continue a pattern 
already set, rather than add to problems by trying 
to build an adequate research organization as well. 
Third, the other institutions and agencies by this 
time would have developed vested research interests, 
and would tend to encourage a situation in which 
they, not the Service, carried out the research 
required. A final reason would be economy. As 
many of the projects, especially in the early years 
were carried out by organizations with their own 
funding sources, the Service received the benefits 
of research gratis. 

Whatever the reasons, the result, over the years, 
has been for administrations to emphasize construc- 
tion, planning, concessions, visitor services, often 
at the expense of programs and organizations to 
study and serve the resources to be preserved. 2 
With the influence of the large natural parks be- 
coming, from the outset, the overwhelming concern 
of Service management, historic and prehistoric 
sites and research needs assumed at best a second- 
ary status. The National Parks Act spoke of con- 
serving "the natural and historic objects, and 
the wildlife therein ..." Historically, however, 
the emphasis of Service programs has been on the 
natural objects and the wildlife, and even the 
former have not received the research support 
natural scientists believe is their due in pre- 
serving the Nation's irreplaceable treasures. 

Although the effort has been too small in my 
opinion, archeological research, at one funding 
level or another, has been a continuing part of 
Service activity and significant contributions 
have been made to the body of knowledge husbanded 
by our Nation's anthropologists. The emphasis our 

This is not to say that maintenance of the re- 
sources has not been a concern. Very good ruins 
stabilization has been done almost from the begin- 
ning. I merely express the view that our efforts 
have suffered from too limited funding; less than 
fully trained personnel, and lack of enthusiasm 
for research, per se, where greater concern would 
have enhanced visitor experience and more enlight- 
ened resource management. 


efforts have taken has reflected the intellectual 
biases of several generations of scholars in 
American prehistory. Early emphasis was on collec- 
tion of objects — specimens for study and for exhibit 
purposes. Later, excavations in some Service areas 
contributed data to local cultural sequences. Today, 
studies focus on Man and his interaction with the 
natural environment, and currently, there is concern 
with "conservation archeology," i.e., the conserva- 
tion of sites for the future, if alternatives exist, 
rather than excavation now. 

Many of the earliest Service studies were brought 
about by the need to preserve southwestern ruins 
for exhibit to the public; to clear them and sta- 
bilize their walls for purposes of preservation 
and visitor safety. This was the intent of Fewkes ' 
(1909, 1911) early work at Mesa Verde. 

In the early years, particularly through the 
1930' s, major projects of large scope were a rarity. 
In part this reflected the state of the art of the 
time. It was really only after World War II that 
the Service sponsored any large archeological pro- 
jects such as the Wetherill Mesa project. The 
Wetherill Mesa development was an outgrowth of the 
philosophy generated by the Service's Mission 66 
program. Chapin Mesa and its ruin complex were 
suffering from excessive visitor impact. In keep- 
ing with the Mission 66 idea of promoting greater 
dispersal of visitors through parks in an effort 
to reduce impact, the Wetherill Mesa ruins were 
viewed as prime resources for development. It was 
thought such development would offer alternatives 
to the visitors. Although it fell somewhat short 
of its goals, the Wetherill Mesa project was a 
landmark in Service archeological work. It was 
administered and directed by Service personnel. 
Excellent physical facilities were provided, and 
it was adequately staffed on a multidisciplinary 
basis. Only the current Chaco research project 
is comparable in the field of prehistory. Directed 
by Douglas Osborne between the years 1958 and 1963, 
the Wetherwill Mesa project resulted in the prep- 
aration of several major manuscripts, some of which 
are yet to be printed. 

Historical archeology in the Service has taken 
an evolutionary course somewhat separate from that 
of prehistory, although the personnel employed have 
shifted back and forth between the two fields. In 
the 1930' s the Service performed, or sponsored, 
historical archeological projects (e.g., Ocmulgee 
Trading Post and Fort Sisseton, South Dakota) . But 
the effort that dramatically brought the attention 
of scholars to the contributions archeology could 
make to the study of American history, and to plan- 
ning and developing historic sites, was J. C. 
Harrington's excavation at Fort Necessity, Pennsyl- 
vania, in 1952 and 1953. Here Harrington fully 
demonstrated to historical scholars the validity 
of applying archeological field investigations to 
the solution of a historical problem. His work 
showed that previous assumptions concerning the 
location, size, and shape of George Washington's 
fort had been erroneous. He also extracted rather 
detailed information on the original fort struc- 
ture, including the stockade and outer entrench- 
ments, permitting an accurate reconstruction in 
place of the incorrect and conjectural development 
that preceded his work (Harrington 1957) . 

The Fort Necessity study was pr 
ed by several projects of lesser s 
of areas: Theodore Roosevelt NMP, 
and Custer Battlefield, to name th 
major effort in this period was co 
L. Cotter at Jamestown. Beginning 
team of investigators assembled by 
fully the Jamestown site, uncoveri 
exhibitable features as the Glass 
Jamestown project served to establ 
archeology solidly as a form of re 

eceded and fol low- 
cope in a variety 

Fort Laramie, 
ree. A second 
nducted by John 
in 1954, a 
Cotter studied 
ng such eminently 
Furnace. The 
ish historical 
search to be 

applied to the solution of historical problems, 
and Cotter's (1958) report is a classic in litera- 
ture of this field. Jamestown was followed in rap- 
id succession by work at Fort Frederica, Appomattox, 
Cumberland Gap, and Independence. Cotter super- 
vised the Independence project, and here again 
applied the type of team effort that served him 
so well at Jamestown. 

While historical archeological studies were 
burgeoning in the East, major historical studies 
were also initiated in some Western parks, including 
excavations at Tumacacori, Arizona; Fort Union, 
New Mexico; and Bent's Old Fort, Colorado. The 
work at Tumacacori was carried out by the late 
Paul Beaubien (1934 and 1935) and was probably the 
first historical excavation in a Western field 
area. One off-shoot of it was the later study, 
by Louis R. Caywood, of Spanish Majolica ceramics. 
Work at Fort Union was primarily a stabilization 
project, but excavation of various features in 
connection with the stabilization work furnished 
a rich body of information on life at the fort, 
and on the function of fort structures. Some 
minor historical archeological studies were done 
in the Midwest Region in the 1930's, but as at 
Fort Union, major work of this kind did not occur 
until after World War II. I believe that the post 
World War II salvage archeological program did 
much to advance the concept of archeological 
reseach as a contributor to historical programs. 
Many important historic sites were excavated in 
the course of reservoir salvage work, and this 
did have some influence on in-park work as well. 
The excavation of Bent's Old Fort, in the Midwest 
Region, was an interpretive necessity. When the 
area was authorized in 1960, no features of the 
fort survived above ground. It was evident that 
little could be done at the site without a full 
archeological study. In 1963, as Regional Archeol- 
ogist for the Midwest Region, I participated in 
the Bent's Fort Master Plan effort. It was my ex- 
press duty to develop plans for complete excava- 
tion of the fort. In the fall of 1963, Jackson 
W. Moore, Jr. was moved to the area to direct the 
excavations. He pursued this work between 1963 
and 1966, and prepared a report, since published 
by the Colorado State Historical Society (Moore, 
1973) . 

Service concern with archeological studies on 
lands other than those it administers also dates 
from the 1930 's. However, major activity of this 
kind, again, followed World War II. The work of 
the 1930 's resulted from Service involvement in 
Civilian Conservation Corps programs, and from 
projects funded by the Work Projects Administration. 
Following World War II, from 1946 until the present, 
the Service became involved in an archeological 
salvage effort of proportions unparalleled in the 
history of the science. Archeologists knew, be- 
ginning as early as 1944, that the Corps of Engin- 
eers and the Bureau of Reclamation had plans for 
extensive water control projects in thepost-war era. 
In that year, the President of the Society for 
American Archeology appointed a planning committee. 
The committee reviewed the plans of the major con- 
struction agencies as well as the previous Federal 
development related and archeological efforts, such 
as those of the Tennessee Valley Authority, Civilian 
Conservation Corps, and Work Projects Adminsitra- 
tion. As a result of this committee's study, a 
Committee for the Recovery of Archeological Remains 
(CRAR) was established. CRAR was to advise the 
government on the emergency need, level of effort 
required, and trends that archeological work should 
take. The original Committee for the Recovery of 
Archeological Remains was composed of the follow- 
ing members: Williams D. Webb (Chairman), Fred- 
erick Johnson (Secretary) , John Otis Brew and A. 
V. Kidder. William Duncan Strong was appointed as 
a liaison member for the National Research Council, 
and Frank H. H. Roberts, Jr. was liaison member for 


the Smithsonian Institution. The Committee met 
annually in Washington with officials of the 
National Park Service; representatives of Federal 
construction agencies, the Smithsonian Institution, 
and the Secretary of the Interior. During the 
field seasons, individual members of the Committee 
made inspection trips to areas where fieldwork was 
in progress. They also met with key members of 
Congress concerning program needs. 

Through the legal authority mandated under the 
Historic Sites Act of 1935, the Service and the 
Bureau of Reclamation became the agencies respon- 
sible for funding the salvage program. I was once 
told by the late John Corbett (then Chief Archeol- 
ogist for the National Park Service) , that consid- 
eration was given to requesting that the National 
Park Service accept full responsibility for admin- 
istration and execution of the program. At the 
time, the Service Directorate had no interest in 
staffing for a major archeological research effort; 
hence, an arrangement was made whereby the Smithson- 
ian Institution would carry out the actual fieldwork. 
This work began in 1946. Later, as the funding 
level and workload increased, the Service began 
to negotiate contracts and engage the assistance 
of universities. The effort grew to proportions 
unforeseen by anyone who contemplated the program 
at its outset. 

The year 1946 saw, too, the appearance in the 
Washington Office of John M. Corbett, brought in 
from the field to administer the Service's part in 
the salvage program. Corbett, as Chief Archeologist, 
was to have a profound impact on the field of 
American archeology generally, as the scope of 
Federal involvement increased. The salvage program 
was administered by Corbett, who was also respon- 
sible for studies in National Park areas. Thus 
Corbett 's influence on Federal archeology became 
immense. John was a man of vast energy and im- 
posing appearance. From the beginning, he saw 
clearly the needs of the Service in archeology. 
He gave his career and his life, in effect, to see- 
ing that these needs were filled. Few programs 
involving American archeology would be what they 
are today had it not been for the efforts of this 
man. The Service, and the field of American arche- 
ology owe John Maxwell Corbett a great debt, one 
that has been too little acknowledged. 

As was pointed out above, the salvage program 
was administered by the Chief Archeologist, who was 
also responsible for studies in National Park areas. 
This had a profound impact on in-park studies in 
many ways. Obvious and first was the impact on 
funding. The salvage program demanded the lion's 
share of the monies involved — but it also had a 
positive impact, in that concepts, theories, and 
techniques developed in the emergency atmosphere of 
salvage came to be applied to problems in park areas 
or incorporated into park interpretive programs. 
Nevertheless, salvage was "where the action was," 
and I recall vividly my first years as Regional 
Archeologist, wherein I found it necessary to beg, 
borrow, and sometimes seemingly steal funds to 
launch sorely needed projects in the parks of the 
Midwest Region where there had been sad neglect of 
necessary surveys for archeological base map (i.e., 
Resource Basic data) purposes. We hardly knew, in 
those years, what our resources were in the parks, 
and we certainly were in no position to study, con- 
serve, and protect them except as the i.eed arose 
in an emergency. Usually, under such conditions, 
we were too late and badly underfunded to be effec- 

As indicated in the introductory section of this 
paper, stabilization of ruins was a prime concern 
early in the Service's archeology program. Even 
before there was a National Park Service, there was 
a Ruins Stabilization program. The Division of 

Cultural Properties Conservation in the Western 
Archeological Center, is an outgrowth of the 
earlier Ruins Stabilization Team. In the early 
years, ruins stabilization was performed on a 
rather informal basis. In the 1930 's Charlie 
Steen (later Regional Archeologist of the Southwest 
Region) and other archeologists moved about the 
Southwest stabilizing numerous ruins. The late 
R. Gordon Vivian was the preeminent figure in ruins 
stabilization, however. From the 1930 's until his 
death in 1966, Gordon, with the standards he set, 
influenced the stabilization process throughout 
the National Park System. I will not go into the 
details of the development of Gordon's program here. 
The subject is well treated elsewhere by Pierson 
(1956:80-83). Gordon centered his activities at 
Chaco Canyon, but as the years progressed and the 
influence of his work spread to other Southwestern 
areas, pressure mounted for Gordon to move his 
headquarters from Chaco to Globe, Arizona, a move 
unpopular with Gordon because of his attachment 
to Chaco. Gordon resisted this effort successfully 
for some time, but eventually, when the Southwest 
Archeological Center was established it became 
necessary to move the Ruins Stabilization Unit 
into the new research-oriented organization. Even 
then, Gordon personally avoided Globe for a period 
of time. 

For many years, from the time there were 
Regional Offices, responsibility for Park archeol- 
ogical studies was the duty of the Regional Arche- 
ologist. Of necessity, performance of the actual 
work was not centered in one single spot, and speci- 
fic central physical facilities were not existent, 
unless one considers Vivian's headquarters at 
Chaco to have been such. There was constant talk 
of one area or another functioning as a center for 
archeological studies. I have not examined all 
of the legislation creating Service areas, where 
prehistory is a major theme, but it is my impres- 
sion that the concept of such areas functioning 
as "centers" is repetitive. This is the case for 
Effigy Mounds and Ocmulgee National Monuments. In 
no case with which I am familiar, however, has the 
idea been successful. Effigy Mounds certainly does 
not function as a research center, and during the 
short period of time when Ocmulgee did, there was 
constant friction between the area staff and those 
attached to the Southeast Archeological Center. 

The creation of the Southwest Archeological 
Center at Globe, Arizona, in 1954, was not an event 
with obvious portent for the future. At that time, 
when one thought of Service archeology, one still 
thought of the Southwest, forgetting the existence 
of the few scattered areas commemorating prehistory 
in the East. It seemed proper that there should 
be a research center serving this region, but no 
one would have thought of such for other parts of 
the United States. A good idea usually does not 
die, and other centers eventually were born. The 
Southwest Archeological Center at Globe occupied 
property which had an august history in the field 
of American archeology. The old Gila Pueblo had 
been headquarters of Harold Gladwin's research or- 
ganization, and had therefore known the presence 
of figures impressive in the study of American pre- 
history. It was fit that the tradition should be 
carried on by the Service. The second Center, 
headquartered at Ocmulgee, was also appropriately 
located from the standpoint of continuing a proud 
tradition. Numerous people eminent in Southeastern 
prehistory had worked in the basement of the large 
Ocmulgee Visitor Center, and major projects had 
been directed from this base. Again, it was appro- 
priate that the Southeast Archeological Center 
should occupy this space. In a different way, the 
same could be said of the Midwest Archeological 
Center when it came into being. The Midwest Center 
evolved from the earlier Smithsonian Institution, 
River Basin Surveys office, created in 1946 to 


serve the archeological salvage program. The River 
Basin Surveys had performed prodigies in Plains re- 
search. It became the Midwest ARcheological Center 
as an artifact of shifting Federal agency philoso- 
phies and program emphases. There seems to have 
been a feeling within the Smithsonian Institution 
that River Basin Surveys was no longer compatible 
with the Smithsonian's mission. In 1968 the 
Secretary of the Smithsonian Institution appointed 
an ad hoc committee to study the River Basin Surveys 
program. The Committee recommended transfer of the 
organization to the National Park Service, since 
virtually all of its funding came from the Service, 
and the program it served was a Service responsi- 
bility. The shift to the Service was made, effec- 
tive July 1, 1969, and John Corbett appointed me 
as the first Chief of the newly created archeologi- 
cal center. Ultimately, both the Southeast and 
Southwest Archeological Centers were moved to 
university campuses: the Southeast Center to Talla- 
hassee, to an association with Florida State, and 
the Southwest Center to Tucson, associated there 
with the University of Arizona. Today the Midwest 
Center maintains an association with the University 
of Nebraska. 

In 1969, also, still another Center was in stages 
of gestation. This was the Chaco Center, currently 
housed partly on the campus of the University of New 
Mexico and partly at Chaco Canyon. Discussions 
had been ongoing concerning the possibility of a 
major research effort at Chaco Canyon involving the 
Service and University of New Mexico. The University 
had a long history of involvement at chaco Canyon. 4 
Again, it was appropriate that the Chaco Center have 
an association with the University. In the fall of 
1969, the writer and Zorro A. Bradley were sent by 
John Corbett to Albuquerque and Chaco to produce 
a prospectus for a Chaco Center, complete with a 
general research outline and multi-year budget for 
the research to be performed. Based on this and 
other documents, an agreement was reached and the 
Chaco Center was created with Robert W. Lister, 
formerly of the University of Colorado, as its Chief. 

Although Chaco is the only Center created to 
serve a specialized function, it has contributed 
broadly to Service programs in prehistorical and 
historical archeology and through its energetic 
application of remote sensing research. Other 
Centers have developed one kind of specialization 
or another. The Southwest Archeological Center 
(now the Western Archeological Center) always led 
in Ruins Stabilization. For a period of time, due 
to the influence of the Bertrand project, the 
Midwest Center was a leader in the conservation of 
historical artifacts. Its research emphasis has 
always centered on the Plains, and currently it 
maintains a general field research orientation 
while it is developing specialized expertise in 
magnetometer surveys. 

The Southeast Center traditionally has main- 
tained a more generalized posture, but recently it 
has begun to focus on manipulation of computerized 
survey data to illuminate park needs. It is also 
placing heavy emphasis on littoral archeological 
studies--an emphasis that is natural and under- 
standable considering the large number of coastal 
areas with which it is concerned. 

The evolution of one additional phase of Service 
archeological endeavor must be briefly outlined. 
Archeological contributions to the planning process 
have always been a part of Service policy. For the 
most part, in earlier years, this contribution was 

Between che years 1929-47, the University was 
heavily involved in research at Chaco Canyon. For 
a period of years Chaco headquartered the Univer- 
sity's field school in archeology. 

sought only where prehistory was obviously in- 
volved, which meant, usually, that it was sought 
only for prehistoric areas. Time and the appear- 
ance of new legislation has modified this original 
simplistic approach. As we have obtained more 
detailed information on the location of archeolo- 
gical resources in Service areas, and as conform- 
ity to the newer legislation has improved, arche- 
ology has come to play a larger role. Once the 
offices of planning and construction within the 
Service drew upon the Regional Archeologists for 
review of their draft plans or completed plans. 
As the Service was being reorganized in 1971-72, 
archeologists were assigned to the staffs of the 
two Service Centers that then existed (i.e., 
Washington and San Francisco). Paul J. F. 
Schumacher held this position in San Francisco, 
and John L. Cotter functioned in this position for 
Washington, although he was located in Philadelphia. 
With the merging of these two offices into the 
single Denver Service Center, both of these men 
elected to retire, leaving a void in the area of 
archeological expertise and coordination. Late in 
1972, I transferred from the Midwest Archeological 
Center to become the archeologist on the Historic 
Preservation Team of the Denver Service Center. 
Since that time, the Service Center has had con- 
siderable involvement in archeological work, re- 
viewing programs, plans, and projects to provide 
adequate archeological contributions to the work 
underway. The archeological staff of the Service 
Center has also provided researchers to carry out 
limited projects in the field. On a scale some- 
what more modest than that of the archeological 
centers, it has made significant contributions to 
research in the Parks. In one case its contribu- 
tion was considerable, for virtually all of the 
archeological work connected with planning, devel- 
opment and restoration work for the Bicentennial 
of the American Revolution was carried cut by the 
Service Center archeological staff, with the re- 
search performed either through contracts, or, in 
some cases by Service Center archeologists. 

The history presented on the preceding pages 
is, as I emphasized at the outset, general and 
spotty. I believe, though, that it reflects with 
some accuracy the trend and spirit of the activity 
within the Service. For many years the National 
Park Service led all agencies (federal, state, and 
local) in management of cultural resources. Whether 
this is still true today is difficult for me to 
judge. I am too close to the shifting policies 
and budget emphases to be an objective observer. 
I feel that, over all, given the limitations under 
which the program has functioned, the Service has 
done well. In fine, I think our situation is com- 
parable to the story told of that day in the 1930 's 
when Dizzy Dean, the late St. Louis Cardinal 
pitcher, was having a bad day on the mound. When 
the manager came out to ask Dean if they should put 
in a relief pitcher, old Diz shifted his quid of 
tobacco and replied with a question, "Who in the 
Hell have you got that's any better?" 


After a lengthy involvement as a researcher and 
administrator with Federal archeological programs, 
I (F. A. Calabrese) cannot help but agree with 
Logan's assertion that there is not another Federal 
agency better prepared to lead in the field of 
cultural resource management. However, the ability 
to maintain this lead is dependent somewhat upon 
a shift in current management attitudes toward the 
priority of consideration of cultural resources. 
Current Federal legislation, promoted by the pro- 
fessional archeological community and not the 
National Park Service, governing the protection of 
cultural resources is sufficient to protect that 
resource within National Park areas only if the 

6 1 1 

intent of the legislation is followed. The legis- 
lation also outlines the Service and other Federal 
agency obligations with respect to management of 
cultural resources. 

I wish now to point out some current resource 
management problems faced by the National Park 
Service (as well as other Federal agencies) and 
outline some alternatives for resolving some of 
these basic problems. 

National Park Service archeological programs are 
in theory governed by both the legislation which 
was established to protect archeological sites and 
the Service's management guidelines for the preser- 
vation of historical and archeological resources. 
The Service still has a dual function with respect 
to cultural resource management. It is responsible 
for assisting other Federal agencies with adhering 
to legislation governing the protection of arche- 
ological resources on Federal lands and it is also 
responsible for protecting and preserving arche- 
ological sites within National Parks. Some people 
are familiar with the legislation which governs pres- 
ervation and protection of the nation's archeologi- 
cal resources. For those who are not, the more im- 
portant aspects of these are outlined below: 

1. The Antiquities Act of 1906 (PL 59-209, 34 
Stat 225) ; the first in a series of protective leg- 
islation for antiquities, requires individuals under- 
taking archeology on Federal lands to seek a permit 
through the Departmental consulting Archeologist, 
Office of Archeology and Historic Preservation be- 
fore proceeding. 

2. The Historic Sites Act of 1935 (PL 74-292, 
49 Stat 666, 16 U.S.C. 461-467) and Historic Pres- 
ervation Act of 1966 (PL 89-665, 80 Stat 915) . The 
Historic Sites Act of 1935 establishes a national 
policy for the preservation of historic remains of 
national significance. The subsequent Historic 
Preservation Act of 1966 provides the legal back- 
bone for protection of remains significant in 
American history, architecture, archeology, and 
culture. The 1966 Historic Preservation Act, 
through expansion of the Landmark program, estab- 
lishes the National Register of Historic Places; 
defines the procedures which must be followed in 
evaluating cultural remains for inclusion on the 
National Register; establishes the Advisory Coun- 
cil for Historic Preservation as well as defines 
the required consideration (Section 106) for ac- 
tions affecting Registered properties. 

3. Executive Order 11593 (1 
oric Preservation Act of 1966 a 
Sites Act of 1935 making it pol 
agency to inventory its land ho 
ological sites, evaluate these 
Register significance and nomin 
considered worthy of inclusion 
Agencies are to administer and 
tural properties in a spirit of 
future generations. 

971) tied the Hist- 
nd the Historic 
icy for each Federal 
ldings, locate arche- 
sites for National 
ate those sites 
on the Register, 
preserve their cul- 
trusteeship for 

4. The Reservoir Salvage Act of 1960 (PL 86-523, 
74 Stat 220) as amended by the Archeology and His- 
toric Preservation Act of 1974 (PL 93-291, 88 Stat 
174) ; are most important to current Federal arche- 
ological programs. The Reservoir Salvage Act pro- 
vides for the preservation of historical and arche- 
ological data which might be lost as a result of 
dam construction. The amending 1974 legislation 
provides for the preservation of historic and arche- 
ological data which might otherwise be lost as the 
result of any alteration to the terrain caused by 
any agency of the United States or by any private 
person or corporation holding a license issued by a 
Federal agency. 

Over the past 70 years it has been the respon- 
sibility of the National Park Service to interpret 

and administer this legislation and to guide and 
assist other Federal agencies with cultural re- 
source management. It has only been in the past 
5 years that other Federal agencies have taken 
cognizance of their own responsibility for these 
mandates. Up to this time the National Park 
Service assumed the responsibility for coordin- 
ating expenditure of Federal funds provided through 
the Department of the Interior for archeological 
salvage excavation prior to implementation of con- 
struction projects as well as the maintenance and 
consideration of the archeological resources with- 
in National Park areas. This was done with insuf- 
ficient funds, usually with only a single profes- 
sional archeologist per region. More importantly 
the energies of the Service's professional arche- 
ologists were more frequently diverted to assist- 
ing other Federal agencies and not toward resolv- 
ing in-park archeological problems. 

It was not until the enactmen 
Environmental Policy Act and Exe 
combined with the review procedu 
Historic Preservation Act of 196 
dent's Advisory Council that Ser 
began to take cognizance of the 
tural resources in park areas, 
the problems we face with respec 
source preservation and manageme 
parks are the result of an unpla 
casual neglect. 

t of the National 
cutive Order 11593, 
re of the National 
6 and the Presi- 
vice administrators 
impact upon cul- 
In effect now, 
t to cultural re- 
nt in our national 
nned program of 


The problems the Midwest Archeological Center 
encountered with respect to site preservation in 
both the Midwest and Rocky Mountain Regions' parks 
reflect the general problems that prevail in our 
parks through the country. Over the course of the 
past several years personnel from the Midwest 
Archeological Center have visited numerous park 
areas and witnessed considerable destruction of 
archeological sites, particularly in the south- 
western cultural area. We presume the problem 
occurs in other parks — especially those lacking 
the staff and resources to protect their cultural 
remains. It is our belief that the problem is 
compounding. Unfortunately, we have neither the 
personnel, the resources nor the authority to com- 
bat this situation. 

The problem of site destruction falls into 
several categories (1) destruction resulting from 
development and development-related activities 
planned by the Denver Service Center or Regional 
Offices (even though loss may be mitigated, sites 
are still gone forever); (2) destruction resulting 
from small construction and day-labor projects 
initiated by park areas and not coordinated with 
the Service archeologists; (3) destruction by unin- 
formed park personnel, some of whom conceptualize 
the cultural resource as an expendable commodity 
unworthy of preservation; (4) unintentional van- 
dalism, the general deterioration and destruction 
resulting from uneducated visitors climbing on 
standing walls, tramping across sites, carrying 
off artifactual materials, digging in mounds, 
picking at rock art, etc., out of curiosity; (5) 
intentional vandalism, the looting, digging and 
intentional destruction of walls, hearth areas, 
mounds, habitation and storage structures, etc., 
done out of malicious mischief or hope of monetary 
gain and (6) general deterioration as a result of 
wear by natural forces. The results and/or obser- 
vations that have been made of these activities 
have been documented and the data are on file at 
the Midwest Archeological Center. 

Two incidents of site destruction, one by day 
labor and a second by a contract/development oper- 
ation, occurred recently in Grand Teton National 
Park. The first resulted in destruction of a 
major and what could have been highly significant 
archeological site at the outlet of Jenny Lake. 
The site was destroyed when heavy equipment was 

utilized to move partially blown-down trees during 
the summer of 1974. The archeologist who first 
noted the site indicated that 18" of archeological 
materials had been removed. Clearly, the trees 
needed to be removed, however, this destruction 
without mitigation could have been avoided had the 
procedures established by NEPA and other protect- 
ing legislation, as well as the Service's historic 
preservation policies, been followed. 

The second incident is the destruction of the 
archeological site on the north end of the Jackson 
Hole Airport runway. This occurred in clear vio- 
lation of our policies, our stated objectives in 
the environmental impact statement and statements 
made to the Wyoming State Historic Preservation 
Officer and Advisory Council on Historic Preserva- 
tion. Most frightening, however, is the precedent 
set by this action. In this case the environmental 
impact statment becomes a meaningless document, a 
paper exercise to soothe environmentalist concerns. 
It leaves the Service open to severe criticism. 

The recent looting of archeological remains at 
Curecanti National Recreation Area by tourists who 
were shown sites by park personnel was noted earlier 
this spring by an archeological crew under the 
direction of Dr. David A. Breternitz. This we be- 
lieve is carrying the purpose of recreation areas 
a bit too far. But it is indicative of the atti- 
tudes on the part of some park personnel for other 
park areas. 

At Glen Canyon National Recreation Area we have 
documented vandalism of archeological remains by 
an employee of a concessionnaire. Further, it has 
been suggested on a national television program 
that Glen Canyon was an area to "search for arrow- 
heads and Indian remains." This gives the impres- 
sion that collection is an acceptable practice 
within our National Park Service areas. 

us to provide adequate coverage given the size of 
our staff, funding and the time-frames generated. 

The following steps must be taken if we are to 
resolve some of these major problems within the 
park areas. 

In addition, a 
graph Site, which 
techniques and ma 
Here, since methy 
site in April 197 
been pecked into 
pictographs have 
This particular s 
tion, is still a 
approximately $2 5 
of this and simil 
protected by freq 
sible and subject 

t Glen Canyon the Davis Pool Picto- 
is being used to test preservation 
terials, has been vandalized. 
1-methacrylate was applied to the 
6, several fist-sized holes have 
the sandstone panel. The white 
also been outlined by charcoal, 
ite, though destined for inunda- 
National Register property and 
,000 is invested in preservation 
ar sites. It is unmarked and un- 
uent patrol. It is easily acces- 
to vandalism. 

We have also witnessed pecking in an attempt to 
remove pictographs. We observed men urinating on 
rock art panels and have noted the use of standing 
walls for other excretory purposes, we have wit- 
nessed scratching, scraping and adding of contem- 
porary symbols, names and other graffiti to rock 
wall faces. 

The types of deterioration hardest to control 
are those occurring through the natural elements 
and/or human consumption by visitation. During a 
recent aerial reconnaissance of Mesa Verde we 
witnessed archeological sites covered by tin sheet- 
ing and metal which had collapsed. Other sites 
are partially exposed and partially protected. 

The Service has initiated steps to resolve 
some of these problems. Specifically, attempts 
have been made to integrate archeological program- 
ming with developments from the initial phases of 
planning, design/development and ultimately, con- 
struction, iiowever, because of the number of arche- 
ological sites and the rapidity at which develop- 
ment sometimes occurs as a result of add-on pack- 
ages and politically sensitive projects (such as 
the Grand Teton airport) , it is most difficult for 

1. Acquaint s 
with the kind and 
within their area 
resources and the 
with the public i 
if any parks now 
Management Polici 
have very differe 
historic areas 
educational train 
by Service archeo 
approach for some 
personnel or fund 

uperintendent and park personnel 

quality of cultural resources 
s, the means of managing these 
problems they will encounter 
n managing these resources. Few, 
have developed Cultural Resource 
es and obviously, recreation areas 
nt preservation problems from 
This could be affected through an 
ing course set up and/or monitored 
logists. We have discussed this 

time, but do not have the 
s to execute such a program. 

2. A second major effort is needed in the pro- 
gramming of sufficient funds for signs, barriers 
and/or other means of protecting pictographs and/ 
or standing structures and ruins. In the Horse- 
shoe Canyon National Park where extremely signi- 
ficant and unique rock sites occur, the use of 
educational signs appears to help with the problem 
of tourist defacement of these National Register 
properties. Signs, obviously, will not deter the 
malicious vandal, but it can eliminate much of the 
vandalism done out of curiosity and ignorance. 

3. Additional funds for inventory work are 
needed so that those sites most endangered can be 
listed and programmed for protection. I must 
stress that we are talking about large dollar 
amounts for the preservation of sites in this 
particular manner. 

4. Funds must be programmed and projects imple- 
mented immediately to stabilize sites at Mesa 
Verde, Glen Canyon, Canyonlands and other south- 
western and peripheral areas. 

5. Additional funds and personnel must be made 
available for the coordination with the Regional 
Offices and the Denver Service Center to monitor 
upcoming development and construction-related 
programs in our national parks. Funds must be pro- 
vided to maintain liaison, hire and maintain on- 
site archeologists and provide the necessary coord- 
ination between the regional offices and field 

6. Additional funds and personnel are needed 

to train preservation specialists for stabilization 
and maintenance of ruins and sites subject to deter- 
ioration by natural forces and heavy visitor use. 

7. Large sums of money are needed for basic 
developmental research for preservation materials 
and innovative projects such as sub-surface site 
location and mapping utilizing remote and direct 
sensing techniques. 

Over the course of the past three years, Service 
archeologists have asked for and stressed the need 
for economic and personnel support to meet the 
cultural resource preservation needs. Within the 
budgetary limitations available, we recognize that 
the Service has provided invaluable support. Admin- 
istrators recognized our problems and needs, but 
faced with other problems feel unable to commit the 
resources necessary for the preservation of cultural 
remains. In other words, for some reason the prior- 
ity of cultural resource protection and archeological 
research generally slips to the bottom of the 
priority list. We must caution now that the toll 
upon that resource is becoming greater by the year. 
With limited staff and funds we are unable to keep 
up with the pressure upon those resources, and it 

r. 2 

is most painful to watch their deterioration. 


It is realized that this is a solution to the 
problem of preservation; to an administrative, not 
a research problem. In addition to protection and 
preservation more and better planned research is 

Inventories to resolve research (as well as man- 
agement) problems must be pursued within our Nation- 
al Parks. Some thorough archeological surveys, ini- 
tiated in the early 1960's, at parks such as Dino- 
saur and Colorado National Monuments were well plan- 
ned and executed. They now provide basic data for 
management as well as developing interpretive data 
and generating archeological information for these 
parks. These surveys were the forerunners to the 
complete inventories now required under Executive 
Order 11593. 

It is incredible that we are only now just com- 
pleting the archeological inventory at Mesa Verde 
National Park and that that inventory was done on a 
little over $10,000 a year for the past eight years. 
That those funds were inadequate to cover a park 
this size is apparent. Many major parks have not 
been inventoried in the Midwest and Rocky Mountain 
and other National Park Service Regions. 

Recapitulating, we have outlined the development 
of archeological programs within and to the Nation- 
al Park Service since the enactment of the Antiqui- 
ties Act of 1906. We have traced the Service's 
involvement in assisting other Federal agencies 
with cultural resource management problems, long 
before most agencies were aware of or willing to 
accept that they had problems. 

The initiation of new 
the development of innova 
toric archeological resea 
al Park Service funding a 
salvage programs and work 
siderably more would have 
the Service archeologists 
agencies during the rapid 
and construction followin 

research techniques and 
tive prehistoric and his- 
rch was stimulated by Nation- 
nd/or guidance through the 
in National Parks. Con- 
been lost were it not for 
assisting other Federal 
periods for development 
g World War II. 

While the Service continues to assist other 
Federal agencies, emphasis has shifted to the more 
pressing problem of the management of archeological 
resources within National Park areas. Despite ex- 
isting legislation, policy and intent of the Nation- 
al Park Service, erosion of the cultural resources 
within the park areas continues. Various types of 
deterioration of the resources are considered. 
Steps to resolve some of these problems are pro- 

ANONYMOUS, n.d. National Park Service Southwest 
Archeological Center, Globe, Arizona. 

Unpublished document in the files of the South- 
west Regional Office, Santa Fe, N. Mex. 

COTTER, JOHN L. 1958. Archeological excavations 
at Jamestown, Virginia, Archeological Research 
Series No. 4. National Park Service, U.S. 
Department of the Interior, Washington, D.C. 

FEWKES, JESSE WALTER. 1909. Antiquities of the 
Mesa Verde National Park — Spruce Tree House, 
Smithsonian Institution Bureau of American 
Ethnology, Bulletin 41. Washington. D.C. 

FEWKES, JESSE WALTER. 1911. Antiquities of the 
Mesa Verde National PArk — Cliff Palace, Smith- 
sonian Institution Bureau of American Ethno- 
logy, Bulletin 51. Washington, D.C. 

HARRINGTON, J. C. 1957. New Light on Washington's 
Fort Necessity: a Report on the Archeological 
Explorations at Fort Necessity National Battle- 
field Site. Eastern National Park and Monument 
Association. Richmond, Va . 

JOHNSON, FREDERICK. 1966. Archeology in an 
Emergency, Science, vol. 152, pp 1592-1597. 

KING, DALE S. 1956. Some Historical Information 
Regarding Gila Pueblo. Unpublished memorandum 
in the files of the Southwest Regional Office, 
Santa Fe, N. Mex. 

LEE, RONALD F. 1970. The Antiquities Act of 1906. 
Office of History and Historic Architecture, 
Eastern Service Center. Washington, D.C. 

LEE, RONALD F. 1972. Family Tree of the National 
Park System: A Chart with Accompanying Text 
Designed to Illustrate the Growth of the Nation- 
al Park System, 1872-1972. Eastern National 
Park and Monument Association. Philadelphia. 

LISTER, ROBERT H. 19 68. Archeology for Laymen 
and Scientists at Mesa Verde, Science, vol. 
169, pp 489-496. 

MOORE, JACKSON W., JR. 1973. Bent's Old Fort: 
an Archeological Study. State Historical So- 
ciety of Colorado. Denver, Colorado. 

PIERSON, LLOYD M. 1956. A History of Chaco Can- 
yon National Mounment. Unpublished ms . in the 
files of the National Park Service. 

WEDEL, WALDO R. 1967. Salvage Archeology in the 
Missouri Basin, Science, vol. 156, pp 589-597. 

In sum, we are suggesting that we ( 
a training program for initiating mana 
personnel into the sensible use, utili 
management of cultural resources; (2) 
a program to continue our survey and i 
park areas to allow proper management; 
provide money and personnel for basic 
al research projects for preservation 
zation; and (4) should develop a sensi 
goal for the use of cultural resources 
tation and for providing basic informa 

1) must develop 
gers and park 
zation and 
must develop 
nventory of 

(3) must 
and stabili- 
ble long-range 

for interpre- 
tion to the 

f, i 


J. Robert Stottlemyer 1 


The Delaware Water Gap National Recreation Area 
(DWGNRA) was established in 1965. It is located 
along the Delaware River between Stroundsburg, 
Pennsylvania, and Port Jervis, New York (Fig. 1) . 
The Delaware River is one of the few remaining 
drainages within the northeastern United States 
not fully utilized. Over the last fifty years there 
has been considerable planning designed to regulate 
water flow within the basin. Congress authorized 
the Tocks Island Lake (TIL) project under the 
Flood Control Act of 1962. This project was to 
consist of an earth and rock fill dam and reser- 
voir which was to provide for flood protection, 
water supply, power generation, and recreation. 
The TIL would raise the river water level 160 ft 
and would include relocation of U.S. 209 on much of 
the Pennsylvania side. The TIL was to have been an 
integral part of the DWGNRA, but its future now 
appears in doubt. 

The recreation area has an authorized acreage 
of 72,000. Approximately one-half of this has 
been acquired. No major developments except for 
a combined headquarters-maintenance facility on the 
Pennsylvania side have been put in by Park Service, 
and none is presently planned. 

The park lies within two major physiographic 
provinces. The entire area was glaciated during 
the Pleistocene Period. As a result, soil pro- 
files are generally poorly developed and very shal- 
low in many areas of the park. The park is with- 
in the oak-chestnut forest region with much of the 
area forested in hemlock-birch especially the 
poorer sites. Much of the river lowland is pres- 
sently farmed under special use permit. The 
counties surrounding the park vary greatly as to 
their economic standing with the counties in New 
Jersey generally better off financially than those 
on the Pennsylvania side. 

In gathering the data for DWGNRA, much time was 
spent in attempting to define specific objectives 
for application of the information. The recogni- 
tion of data packages as a formal component of the 
planning process within the Park Service is rela- 
tively new. De facto, it is still experimental. 
The effort at DWGNRA was only the second undertak- 
ing involving the interfacing of automated data 
handling techniques with geographical and social- 
economic data. In no manner was this an attempt 
to develop a management information system if 
such can even be defined. The sole emphasis was 
on a planning data package. The reason for this 
is the planning component of the decision-making 
process within any agency is generally the simplest 
cycle to project, and decisions can be easily ar- 
rived at concerning appropriate data quality to 
synchronize with the planning cycle. In any exper- 
imental data acquisition and synthesis effort, one 
can ill-afford to initially expend en rgy on infor- 
mation not immediately applicable to the decision- 
making cycle. 

Guidance as to the quality of data needed came 
from a variety of sources. The enabling legisla- 
tion for DWGNRA obligates the Park Service to pre- 
pare a land management plan for those areas suitable 

■"■National Park Service, Mid-Atlantic Regional 
Office, Philadelphia. 

for recreation, 
ural and histori 
multiple use nee 
timbering. Addi 
as the National 
lines what role 
National Park Sy 
nating propertie 
Historic Propert 

areas possessing outstanding nat- 
cal values, and any lands meeting 
ds such as hunting and possibly 
tional ideas came from documents 
Park System Plan (1971) which out- 
individual park areas play in the 
stem. The criteria used in desig- 
s for the National Register of 
ies can also be of assistance. 

The Park Service interprets its 19 
Act and individual park area's enabli 
through administrative policies which 
ed (1975) . Since 1975 the emphasis i 
development and implementation of pol 
class as arrived at through the plann 
This requires as a first step conside 
nature of the resource to be managed, 
tual land classification map is then 
ified by other factors such as social 
conditions especially those of the co 
counties and communities. 

16 Organic 

ng legislation 

are publish- 
s on the 
icies by land 
ing process, 
ration of the 

The even- 
further qual- 

Policies related to resource management and 
visitor use are specifically implemented in a 
General Management Plan which consists of three 
major parts: A General Development Plan, a 
Visitor Use Plan, and a Resource Management Plan. 
It was the General Management Plan for which the 
data package for DWGNRA was designed. 

In summary, the strategy was to review in de- 
tail the park's enabling legislation, the Park 
Service's administrative policies, and the park 
superintendent's management objectives in pre- 
paring a specific list of issues to be addressed 
in the near future with some indication of their 
priority. Then experts primarily from the acad- 
emic world, were consulted to seek out what kinds 
of information were needed to answer such issues 
and where such information might be found. The 
decision to use automated data handling was based 
primarily on the amount of information to be 
gathered, its synthesis, and the need to evaluate 
a great number of possible contingencies in a 
relatively short time span. 


Funding for this project was made available 
in the summer of 1974. A contract for the gather- 
ing and anlysis of regional social-economic data 
and a visitor use profile was let with the Depart- 
ment of Sociology, City University of New York. 
The gathering of resource data within and cotermi- 
nous the park was by contract with the National 
Aeronautics and Space Administration at Bay St. 
Louis, Mississippi, which at the time had a unit 
established to prepare computer-based land use 
data packages. The decision to go with NASA was 
largely determined by the time constraints of the 

For the social-economic analysis, the six 
counties surrounding the DWGNRA were intensively 
examined. Only existing regional information was 
examined in this phase of the analysis which in- 
cluded among others the following items: available 
public services; population income, movement, age 
trends; economic diversity; trends in county 
budgets and their appropriations; and primary in- 





Stroudsburg, P< 


^ *m 







FIGURE 1. Skylab photo of DWGNRA and coterminous lands. 


A visitor use survey was also conducted. The 
objective was to determine present use patterns and 
trends within the park to get some indication of fu- 
ture recreation and facility demands. The survey 
was conducted on a busy late summer weekend. Six 
sites were selected at which 1000 questionnaires 
were distributed to one adult member of each travel- 
ing group. The questionnaires were returned by 
mail. The questionnaire was approved by the Of- 
fice of Management and Budget. Three hundred and 
seventy-two were filled out and returned. 

In evaluating resource characteristics within 
and coterminous the park, the objective was to de- 
termine resource suitability for a wide variety of 
possible park uses. The overall procedure con- 
sisted of the following steps: define the land use 
objectives, collect the basic data and process it, 
basic data digitization, generate base maps and 
overlays, convert polygon data to grid cell, de- 
rive further information from cell data, determine 
cell suitability for variety of potential land uses, 
set up environmental impact analysis, and develop 
land use plan alternatives with assessments. 

The software program used in the land suitability 
analysis was the Harvard Grid (SYNMAP) significantly 
modified by the School of Forest Resources at Penn 
State University. To insure consideration of all ap- 
plicable data, a matrix was generated showing the 
categories of data on the abscissa with geographi- 
cal areas down to the county level or the ordinate. 
Data were collected primarily from the park and 
Federal, state, and local agencies. All informa- 
tion except topographic (elevation, slope, aspect) 
was obtained in the form of maps. Topographic in- 
formation was purchased from the Defense Mapping 
Service on tapes in a grid format. Maps were dig- 
itized as polygons requiring that all map infor- 
mation be identified by alpha-numeric coding. A 
Numonics digitizer was used. The tapes containing 
the data from each source map were then plotted at 
the scale of the original source map to check 
accuracy. All maps digitized were then converted 
to the same scale (1:62,000). All data were then 
merged onto a single data file on tape. 

Often source maps have errors such as distortion. 
Thus the previous steps can be time consuming and 
expensive. Considerable time was spent on this 
stage to ensure map accuracy so mismatch of infor- 
mation did not occur when the information from the 
varying sources was combined. With the correction 
of errors, the digital tape was ready for final 
overlay plotting at which time titles, legends and 
any other map information thought appropriate was 
combined on the overlay. Overlays were produced 
on a high-speed Xynetics plotter and the follow- 
ing information was displayed: maximum topography 
line, flood prone areas, drainage, TIL reservoir 
water line, DWGNRA boundary, land ownership, ele- 
vation, aspect, slope, vegetation, soil type, soil 
depth, geology, surface water, present land use, 
transportation, utilities, historic structures, 
archaeological sites, and critical areas. 

This in 
hectare gr 
of resolut 
stance in 
be made, 
checked ou 
cisely by 
above info 
park bound 

In the 
relevant 1 

formation was then converted to a one 
id cell format. The one hectare level 
ion was thought sufficient in this in- 
view of the nature of the decisions to 
The accuracy of the grid conversion was 
t by line printout display ^nd more pre- 
Ramtek (cathode) dxsplay. All of the 
rmation was available for land within the 
ary. Soils and vegetation data were not 
obtained for the coterminous 250,000 

determination of resource suitability or 
ness for each grid cell, the following 
was used. Data variables pertinent or 
n evaluating the specific land use were 

selected. The relative weight or importance of 
the various levels within each variable were deter- 
mined. Third, an overall multiplier to a parti- 
cular data variable was assigned if it was felt 
one variable was quantifiably more important than 
another in the evaluation of resource suitability 
for a particular land use. 

Examples of the types of land use evaluated at 
DWGNRA were trail hiking, intensive camping, swim- 
ming, high intensity recreation, access roads, 
fishing, hunting, farming, and boating. 

An additional software procedure was tested to 
examine the importance of adjacent cell charac- 
teristics and their effect on the primary cell 
suitability for a particular use. The objective 
was to determine if overall suitability of park 
lands for a particular activity is significantly 
altered when the characteristics of coterminous 
cell layers are considered. 

Several software programs were also tested in 
an attempt to utilize and evaluate the impact of 
planning contingencies. The three procedures 
sampled involved considerable differences in com- 
puter programming and computer time and none ap- 
proached the ideal. The methods experimented with 
were the matrix approach, the inverse of cell suit- 
ability, and determining the mean suitability of 
all cells involved in a particular plan alterna- 
tive. Space does not permit an elaboration of 
the specifics of these three procedures. However, 
the matrix approach is by far the most complex and 
most expensive. 


The project, with some exceptions, was completed 
on time. There were some procedural problems which 
are discussed in the following section. To indi- 
cate how the information results will be applied, 
highlights of the data analyses follow. 

The results of the social-economic evaluations 
were the most revealing in terms of providing new 
information contrasting with popular belief. 
First, there is a great divergency in the ability 
of the coterminous counties to absorb the addition- 
al costs of services that will be required should 
visitor use of DWGNRA significantly increase. In 
general, counties on the Pennsylvania side are 
less equipped to handle any increased burden of 
additional public services. New Jersey counties 
have a younger and more diversified population with 
heavy reliance upon the New York City metropoli- 
tan area for a source of income. Recent changes 
in the quality of social services and present 
strengths regarding police, fire, transportation, 
and water services will allow these counties to 
absorb the costs of expanded social services more 
easily. Conversely, the provision of other types 
of services such as intensive camping is better 
provided relative to demand on the Pennsylvania 

The type of present visitors and their demands 
on existing services also varied from popular be- 
lief (Robison 1976) . Fifty-four percent had in- 
comes in excess of $16,000/yr. Most come from the 
urban New York City area. Families with incomes 
under $10,000 or students make up a pronounced 
minority of present park visitation. Many visitors 
own second homes in the DWGNRA vicinity. Most 
visitor use consists of a one-day outing or week- 
end outing, and relatively little money is spent 
locally for meals or accommodations in view of the 
relative economic strength of present park visitors. 
Likewise, demands on park resources and services 
are mostly non-consumptive. Chief activities in- 
clude hiking, photography, bird-watching, and 
antique shopping. Present visitors would rather 


see little change in park-provided facilities re- 
questing only a slight improvement in basic neces- 
sities and improved access to park information. 

These resul 
ning qualifier 
park's general 
results will s 
vices the park 
plied by cote 
guide the sequ 
ments so impac 
gated or at le 

ts will be applied as overall plan- 
s in the ongoing preparation of the 

management plan. Primarily these 
erve to determine what goods and ser- 

will provide and what will be sup- 
rminous communities, and also to 
encing of any eventual park develop- 
ts to coterminous areas can be miti- 
ast minimized. 

The final determinations of resource suitability 
also revealed some interesting results. Although 
the area is glaciated, a surprisingly large amount 
of park land (about 4500 hectares) meets average 
engineering criteria needed for locating develop- 
ment. If one examines present visitor use levels 
per developed acre in existing national parks 
(which reflect the average intensity Of park devel- 
opment along with the willingness of Congress to 
appropriate funding for development) , one finds a 
daily visitor use density of from 1-6 people per 
acre per day (Sudia et al. 1975) . If one uses 
the mean visitor use density per developed acre 
of park land multiplied by the number of acres 
potentially suitable for development at DWGNRA, 
it is clear that suitability of the resource with- 
in DWGNRA will not be a limiting factor to park 

There has been significant ground truthing done 
to confirm the results of the land suitability 
analyses. Areas indicated to have high suitability 
for development have been examined. Likewise, many 
areas shown to have very limited suitability have 
been examined to confirm the indicated limiting 
factor. One present visitor use demand which will 
probably be expanded is the use of trails. Many 
trails have been examined on the ground along with 
the suitability maps for locating trails to deter- 
mine what difficulties might exist if present trails 
were expanded in use. 

An interesting application of the suitability 
data came in locating a site for a septic tank 
for the Bushkill headquarters and maintenance area. 
The original contract anticipated such treatment 
but a site could not be located and an attempt was 
made to amend the contract upward for a $160,000 
tertiary plant. Querying the suitability maps lo- 
cated four potential sites within reasonable dis- 
tance and one was selected. 

An initial draft of a land classification map 
is being prepared which will divide the park into 
those lands suitable for development, possessing 
outstanding natural value, historical value, and 
those remaining lands which might serve multiple 
uses. Within each of these land classes there 
will be subzones based upon resource suitability. 
In the final analysis, the planner will not be re- 
stricted to classify the land strictly according 
to resource suitability. Other factors as social- 
economic conditions will qualify the land classifi- 
cation imposed. But regardless of the eventual land 
classification scheme, the planners will know the 
relative constraints and attributes of the resource 
they may be using. The next step will be to inte- 
grate those quantifiable aspects of the social- 
economic survey with the geographical data in an 
attempt to assist planners developing a final land 
classification map. 

Prior to completion of planning for DWGNRA, 
which will probably occur in the early summer of 
1977, there will be further work done on develop- 
ing a better model for the evaluation of the en- 
vironmental impact of planning alternatives. Inte- 
grated with this will be an attempt to better quant- 


ify long-term maintenance costs associated with 
locating similar developments on sites of varying 
suitability. This may be the most important 
product resulting from the generation of soph- 
isticated data packages for park planning. For in 
all probability it is the unanticipated main- 
tenance contingency which creates the most havoc 
for a resource management agency. The inability 
to objectively quantify in the long-term the oper- 
ational costs of a park and related maintenance 
emergencies is what uses up funding originally in- 
tended for long-term projects such as research 
and resource management. 

Also to be completed and utilized in the on- 
going planning effort will be a system approach 
in determining the capacity of the developed com- 
ponents within the park so visitor distribution 
and flow will take into account capacities for the 
developments leading into and away from a given 


The preparation of data packages for the nation- 
al parks has met only mild enthusiasm. There are 
many reasons for this but certainly three stand 
out. They are (1) the complexity of preparing 
such information bases, (2) the difficulty in ap- 
plication of the data to the decision-making pro- 
cess, and (3) the cost of such undertakings. 

In the preparation of information bases, dif- 
ficult decisions must be made very early in the 
process. The system employed must recognize the 
end use of the information gathered. If the in- 
formation is to be used in resource management, 
who really controls this function in the organiza- 
tion and how will the information be used and 
what information is needed? Throughout the pro- 
cess is the danger of over-emphasis of technology 
in information handling. Hardware and software 
are much advanced compared to the realistic needs 
of most federal agencies. There is a real danger 
in expending funds just to gather information es- 
pecially geographical information. 

There are also complex questions related to 
standards for data acquisition. Our experience 
in the efforts at DWGNRA has shown the lack of 
adherence to standard procedures in mapping and 
listing of information is extremely expensive to 
correct when combining information from varying 
sources. This problem the information synthesizer 
has little control over. A decision then has to 
be made whether to take the time to correct source 
information or remove it from the list of desired 
data. Another early question to answer is whether 
to use automated data handling techniques. Ideally, 
the question could be answered by optimizing eco- 
nomic and time factors. However, such information 
is rarely available. The decision must then be 
based primarily on the amount of information to 
be handled and the time in which it must be avail- 
able in a synthesized form. 

But the item most likely to unduly complicate 
the preparation of information bases is the ini- 
tial poor definition of problems to be resolved. 
The lack of specifically itemized objectives and 
the lack of clearly understood end-products in- 
creases the chance of taking an inordinate number 
of tangents in the synthesis of sophisticated 
information bases. 

The second difficulty is perhaps the most cru- 
cial that being how to phase the information 
quantity and quality with the decision-making 
cycle. Early in the preparation of any informa- 
tion base a decision must be reached on whether 
to concentrate on general or specific information. 
That is whether to side in the favor of decision- 

making qualifiers or information to quantify speci- 
fic aspects of problems. At DWGNRA one of the early 
mistakes was not providing sufficient emphasis on 
social-economic qualifiers. Too much emphasis was 
placed upon traditional quantifiable geographic 

The anatomy of the decision-making structure 
of an organization must be known by someone in- 
volved in the preparation of information bases. 
Park management is carried out in a cyclic manner. 
Decisions are made on weekly, annual, multi-year 
cycles and an information base must be so designed 
to provide the right types of information at the 
proper resolution synchronized with the decision- 
making cycle. The anatomy of the decision-making 
cycle also assists in making a decision on when 
to interface computers in data handling. If the 
information needed must be highly mobile, that is 
readily available for the analysis of a variety of 
contingencies, then it is probable automated data 
handling should be employed. Part of the decision- 
making cycle consists of the planning cycle which 
is perhaps the best place to start information 
bases since the types of decisions and their se- 
quence are fairly easy to understand and project. 
The relative straightforward design of planning 
within Park Service was the reason the information 
base at DWGNRA initially concentrated on planning 

But even a full 
the decision-making 
the information pac 
there appears very 
Park Service where 
interfaces with an 
lies upon subordina 
formation system is 
in terms the decis 

and accurate comprehension of 
cycle does not insure use of 
kage generated. At present 
few instances in or outside the 
the decision-maker directly 
information base. Rather he re- 
tes to interrogate whatever in- 
available and provide results 
ion- maker can comprehend. 

But in the end, it must be recognized the 
basic value of information bases has to be, in some 
manner, economic. Somebody has to make a decision 
on how much should be expended for additional in- 
formation. In short, the cost of an information 
system is going to depend on what needs to be known 
and how well does it need to be known. Economic 
exchange values, that is cost-benefit alalysis, 
should probably not be applied to information 
bases. The reason for this is obvious. Many 
factors cannot be quantified in such terms. But 
most decision-makers, both within the federal 
government and private industry, do not seem to 
be using cost-benefit analyses as a justification 
for additional data acquisition, but rather the 
utility value of information bases. That is, top 
administration is willing to pay for information 
because it improves their confidence in making the 
correct decision. 


ANON. 1972. Part two of the National Park System 
Plan — Natural History. NPS, Washington, D.C. 
140 p. 

ANON. 1975. Administrative Policies. National Park 
Service, Washington, D.C. 130 p. 

ROBISON, L. K. 1976. Delaware Water Gap National 
Recreation Area: A Visitor Use Survey. Unpub. 
Report submitted to Chief Scientist, Mid- 
Atlantic Region, NPS. 15 p. 

SUDIA, T. W., and J. M. SIMPSON. 1973. Recreation- 
al carrying capacity of the national parks. 
NPS Guideline (3):25-34. 

The third major diff 
of information bases is 
information bases is up 
quires a reasonably lar 
funding early in planni 
many other land managin 
used to spending large 
especially in the early 
information packages ca 
cost nearly $1.0 per ac 
within the park boundar 
the extensive analysis 
acquisition of geograph 
grid format can reduce 
quoted by the U.S. Fore 
an acre. 

iculty in selling the value 
their cost. The cost of 
front. That is, it re- 
ge initial investment of 
ng. The Park Service, like 
g agencies, has not been 
sums of money in planning 

stages. The cost for such 
n be expensive. The DWGNRA 
re for the intensive survey 
ies, and $0.25 per acre for 
of coterminous lands. The 
ical information only in a 
these costs with figures 
st Service as low as $0.10 


Warren F. Steenbergh and Charles H. Lowe 1 


As a direct result of the criteria used in the 
selection of natural features and phenomena for 
designation as national parks and national monu- 
ments, the National Park Service has inadvertently 
acquired unique problems with respect to maintain- 
ing the living natural resources assigned to its 
custody. This is especially true with regard to 
those National Park Service areas that were se- 
lected because of the unique qualities of specific 
plant populations. 

The very qualities that made these plant popu- 
lations attractive, unique or superlative exam- 
ples of their kind, were the unrecognized 
expression of exceptional conditions, i.e., 
atypical environments within the overall species 
range in space or time. Thus the National Park 
Service has been entrusted with the care of plant 
populations that are the resulting relicts of 
past climatic conditions, such as redwoods, and 
sequoias, or populations that are situated near 
the critical geographic limits of the species 
distribution such as saguaros. 

Instability is a primary characteristic of 
such populations responding - often in a dramatic 
manner - to a constantly changing climatic environ- 
ment. Such is the prevailing condition and the 
"problem" of the saguaro population at Saguaro 
National Monument and elsewhere along the margins 
of its distribution in Arizona and northern 

The saguaro giant cactus is a drought-adapted, 
cold-tolerant, warm-desert species that occurs 
throughout the length and breadth of most of the 
Sonoran Desert (see Shreve 1951; Shreve and 
Wiggins 1964; Benson 1969; Hastings et al. 1972). 
Within that distribution it grows primarily on the 
coarse soils of south-facing rocky slopes and 
adjoining bajadas^ and occurs sparsely- or not at 
all on north-facing slopes or in the poorly 
aerated fine alluvial soil on the floor of the 
broad valleys that separate the characteristically 
discontinuous mountain ranges of the Basin-Range 
Province . 

At the southern limits of its distribution in 
southern Sonora, small disjunct populations occur 
on the basalt outcrops of hills. Disjunct popu- 
lations also occur on rocky south-facing slopes 
along the eastern and northern limits of its dis- 
tribution in Sonora and Arizona. The Gulf of 
California marks the western limits of its dis- 
tribution in Sonora. In southwestern Arizona and 
at the westernmost limits of its distribution in 
southeastern California within a few miles of the 

National Park Service, Cooperative National 
Park Resources Studies Unit, University of Arizona, 
Tucson 85721, and Department of Biological 
Sciences, University of Arizona, Tucson 85721, 
A bajada is the broad gently sloping alluvial 

plain extending from the base of the mountains to 
the valley floor. 

Colorado River near Blythe, sparse populations 
are directly associated with areas of concentrated 
runoff. The metropolis of the species population 
near the Arizona-Sonora border coincides more-or- 
less with the geographic center of its distribu- 
tion. The saguaro population at Saguaro National 
Monument (east) occurs at the eastern, cold- 
limited margin of the species' distribution in 

With a maximum expected life-span twice our 
own, living saguaros have survived climatic 
events that long pre-date our recollection or 
knowledge. The species has survived through con- 
tinuing environmental changes and has evolved 
during millions of years in environments that we 
have not experienced, and can but vaguely compre- 
hend. However, examination of the question of 
what has happened, what is happening, and what 
will happen to the saguaro — at Saguaro National 
Monument and elsewhere throughout the range of 
its distribution in the Sonoran Desert — requires 
that perspective. It is within the framework and 
perspective of the evolutionary process that we 
explore the ecology of the saguaro. 

ce has long been con- 
cause of dramatic 
ssly altered the struc- 
s in certain portions 


In the Cactus 

The National Park Servi 
cerned with the nature and 
fluctuations that have gro 
ture of saguaro population 
of Saguaro National Monume 
Forest area of the origina 
ment, the once spectacular 
saguaros has dwindled in a 
impressive population of s 
dying old individuals. 

More important to the present and future 
condition of this population, however, is the 

1 Saguaro National Monu- 
concentration of giant 
few decades to an un- 

parsely scattered and 

Investigations on the saguaro giant cactus 
(Cereus giganteus) Engelm., Carnegiea [Engelm.] 
Britt. and Rose) reported here and elsewhere were 
independently initiated in 1951 by Charles H. Lowe. 
Subsequently the work was continued with the sup- 
port of the National Park Service (and others) in 
recognition of the need for basic knowledge of the 
biology of the species to provide information es- 
sential for ecologically sound management of the 
natural resources of Saguaro National Monument 
and other National Park System areas in southern 
Arizona. Vernacular and scientific names now most 
commonly used are saguaro and sahuaro, and Cereus 
giganteus or Carnegiea gigantea . 

Saguaro National Monument is comprised of two 
separate units situated on opposite sides of the 
Tucson basin. The original, Rincon Mountain Sec- 
tion (east unit) is located approximately 15 miles 
(24.2 km) east of the Tucson city center. The more 
recently established Tucson Mountain Section (west 
unit) is located about 15 miles (24.2 km) north- 
west of Tucson. Shelton (1972) provides a general 
description and account of the natural and human 
history of the Saguaro National Monument. 


relative sparsity of younger plants. That 
condition — the lack of sufficient younger saguaros 
to replace dead and dying older saguaros — insures 
absolutely that the number of large saguaros in 
this population will continue to decline. 

The problem long pre-dates the establishment 
of the original Saguaro National Monument in 1933. 
The predominance of older saguaros , and the lack 
of young individuals in photographs taken more 
than 40 years ago clearly reveal a population 
already many years in trouble. To a greater or 
lesser degree, other saguaro populations within 
Saguaro National Monument--and elsewhere — have 
undergone similar recent fluctuations. 

In recognition of the needs of the National 
Park Service, continuing investigations on the 
ecology of the saguaro are designed to provide 
definitive knowledge of the problem and to develop 
information essential for interpretive and resource 
management programs. Experimental designs and the 
hypotheses that they test are directed specifi- 
cally towards obtaining information on the 
operation of natural selection through climatic, 
and other physical and biotic factors that are 
critically limiting on saguaro populations. This 
report summarizes some of the more important re- 
sults of these investigations contained in 
previous reports (Steenbergh and Lowe 1969; 
Steenbergh 1970, 1972; Steenbergh and Lowe 1976a, 
b) and discusses their relationship to the past, 
present and probable future status of these 
populations . 


The saguaro has a life-span of 150-175 years 
(Shreve 1931) during which it produces on the 
order of 40 million viable seeds. For a population 
to maintain itself or to grow, one of those seeds 
must, within that period, germinate and survive 
approximately 30 years or more to the age when it 
becomes a reproductive member of the population. 
It is hardly surprising, therefore, to discover 
that only a minute percentage of the annual seed 
crop germinates and survives through the first 
year of life to become an established member of 
the population. 

The saguaro blooms and produces an annual crop 
of succulent fruits during the driest period of 
the year. The peak of fruit-ripening and seedfall 
occurs in late June or early July, just prior to 
the arrival of germinating summer rains. 

In the complete life cy 
greatest mortality occurs 
germination (seed) stage, 
between seedfall and germi 
During that period, the ma 
annual seed crop is consum 
birds, mammals and insects 
obligatory seed eaters are 
tion. However, other cons 
omnivorous birds, which pa 
through their digestive tr 
agents of seed dispersal 

cle of the saguaro, the 
during the pre- 
the 1- to 5-week period 
nating monsoon rains, 
jor portion of the 
ed by native animals — 
Seeds consumed by 
destroyed during diges- 
umers , particularly 
ss undamaged seeds 
acts, are important 

The principal germination of saguaro seeds takes 
place in July and August from seeds of the current 
year's crop. Optimum conditions for natural germi- 
nation coincide with the first full development of 
monsoon storms. Then, during the period from 
mid- July to mid-August, germination is normally 
associated with the occurrence of two or more 
rainstorms within a 2- to 5-day period. 

The availability of moisture is the critical 
determinant in the germination process; 

temperature and light requirements are readily 
satisfied within all natural habitats of the 
saguaro. In continuous contact with free water, 
germination takes place in 48 to 72 hours. The 
summer climatic environment of the Tucson area is 
well within the range necessary for saguaro seed 
germination; it exceeds, in germination suit- 
ability, that of the drier, more westerly portions 
of the species' range. At Tucson, depending upon 
the overall intensity of monsoon development, the 
year-to-year suitability of conditions for natural 
germination ranges from poor to near optimum, but 
some natural germination occurs in all years. 

Experimental evidence indicates that fewer than 
1 in 200 seeds that reach a site where germina- 
tion can occur survives to the seedling stage. 
We estimate that the net natural survival to the 
initial stage of seedling establishment is less 
than 1 per 1000 seeds produced. 

For the saguaro, the greatest hazards to 
post-germination survival are associated with the 
pre- juvenile, seedling stage of life. Establish- 
ment is ordinarily reached at an age of 12-14 

The mean life expectancy of the newly emerged 
seedling is on the order of 2-6 weeks. Less 
than 1% of seedlings survive the first year of 
life to become established juvenile plants. 

The tiny, succulent, weakly-rooted saguaro 
seedling is highly vulnerable to destruction by 
a broad variety of abiotic and biotic agents. 
These include freezing, drought, rodents, and 
insects. The relative importance of each factor 
varies with the habitat, the season, and from 

Our data from Saguaro National Monument 
indicate that biotic factors account for the 
greatest number of saguaro deaths during the first 
few weeks of life. However, it is the extremes 
of the climatic environment — freezing and 
drought — that ultimately control the subsequent 
establishment and survival of young saguaro in 
the Tucson area and elsewhere along the northern 
and eastern limits of the plants' range in 
Arizona and Sonora. 

The end of the establishment period and the 
beginning of the juvenile stage of life is marked 
by the advent of monsoon rains of the summer 
following germination. Rapidly increasing in 
size and root development during that second sum- 
mer of growth, the young plant attains a 
substantially increased drought tolerance and an 
increased ability to survive insect attack. 

The life expectancy of a juvenile saguaro 
increases rapidly with age — it outgrows its preda- 
tors, and becomes increasingly able to survive 
the periodic hazards of the abiotic environment. 
In order of their relative importance, freezing, 
rodents and insects are the primary causes of 
death of established juvenile saguaros at 
Saguaro National Monument. With age and increase 
in size of the young plant during the second and 
succeeding years of life, the number and per- 
centage of insect and rodent-caused deaths become 
progressively smaller. Healthy young plants 
greater than 5 cm (2 in) in height are relatively 
immune to destruction by these agents. 

Freezing is the single most important cause 
of young saguaro mortality at Saguaro National 
Monument and elsewhere in the colder portions of 
the species' distribution. The resistance of 


young saguaros to death from freezing increases 
with age-related changes in the size and form of 
the plants. Thus, freezing selectively removes 
the youngest members of the population--seedlings 
and juvenile plants less than 0.46 m (1.5 ft) 

There is little natural mortality of large 
juvenile and young unbranched adult saguaros with 
heights from 0.46 to 3.80 m (1.5-12.5 ft). Plants 
of that size range are relatively resistant to 
freezing, drought, or destruction by animals. 

Vulnerability to climatic factors increases 
with age and height above 3.80 m. Deaths of large 
saguaros are almost entirely the result of the 
action of climatic factors — freezing, windthrow 
and lightning. 

As with juvenile saguaros, freezing is the 
principal cause of death of large saguaros at 
Saguaro National Monument. In this northern por- 
tion of the geographic range of the saguaro, 
freezing selectively removes the youngest (small- 
est) and oldest (largest) members of the plant 
population, and favors the survival of inter- 
mediate age classes. Accordingly, the observed 
decline of saguaro populations at Saguaro National 
Monument, and the unbalanced age structure of 
these populations is primarily the result of re- 
curring catastrophic freezes acting differentially 
upon 1) seedlings and small juvenile plants, 
2) larger juvenile plants, and 3) large adult 
plants . 

Topographic, edaphic, and plant community 
characteristics play a major role in germination, 
establishment and survival of young saguaros. 
Germination is dependent upon the existence of 
physically modified environments produced by 
trees, shrubs, rocks or other shade-producing 
objects. These shaded microhabitats provided mod- 
erated daytime temperatures that are within the 
upper range required for germination, and prolong 
periods of essential high moisture availability 
at the soil surface. 

Establishment and survival of young saguaros 
are likewise dependent upon the physical protec- 
tion provided by close association with other 
vegetation, detritus, and rocks. Acting singly or 
in combination, these moderate the limiting ex- 
tremes of drought and winter cold, and in addition 
reduce the probability of discovery and destruc- 
tion by animals. 

Differential freeze 
young saguaros associa 
intensity and duration 
occurs in topographica 
ferential survival of 
rocky and non-rocky (f 
and south-facing slope 
is highest on south-fa 
facing slopes. In the 
moisture levels that f 
survival in relatively 
set by subsequent free 

caused winter mortality of 
ted with differences in the 

of subfreezing temperatures 
lly different habitats. Dif- 
young saguaros occurs in 
lat) habitats, and on north 
s. Seedling establishment 
cing slopes, lowest on north- 
se habitats, higher soil 
avor pre-winter seedling 

arid environments are off- 
ze-caused winter mortality. 

Growth of the saguaro is dist 
and takes place during coinciden 
tively high temperatures and pla 
moisture. The principal growth 
during July, August and Septembe 
warm summer rains, and ceases sh 
end of those rains. In years wh 
rains provide adequate soil mois 
peratures prevail, a secondary a 
apical growth occurs as early as 

inctly seasonal 
t per'ods of rela- 
nt-available soil 
of saguaros occurs 
r, the period of 
ortly after the 
en late winter 
ture and warm tem- 
nd minor period of 
February and 

March. The total summer growth of the saguaro 
is dependent upon the cumulative duration of 
periods of plant-available soil moisture. The 
annual increment of stem growth, therefore, is 
determined not only by the amount of precipita- 
tion, but in a large measure, by its temporal 
distribution. The annual height growth of young 
saguaros increases exponentially with age and in- 
creased size up to the age of first flowering. 
Then, with the onset of reproductive growth, a 
gradual reversal in the rate of annual height 
growth occurs as energy is increasingly diverted 
into the production of buds, flowers and fruits. 

Apical growth rates provide a basis for 
determining saguaro height-age relationships, thus 
allowing estimates of population age structure. 
Regression analysis of data on the growth of 
seedling and juvenile saguaros in their natural 
habitats at Saguaro National Monument provides, 
for the first time, information on growth and age 
relationships of saguaros during the first critical 
years of life. The annual height growth increment 
for healthy young saguaros increases from a few 
millimeters per year during the earliest years of 
life to approximately 11.6 cm (4.5 in) per year 
during the 33rd year of life, the approximate age 
of first flower production. 

Growth rates of th 
Two-fold or larger di 
growth occur between 
same immediate locali 
Similarly large varia 
ciated with year-to-y 
precipitation and win 
cant differences in g 
different topographic 
general climate. 

e saguaro are highly variable, 
fferences in annual height 
individuals growing in the 
ty and topographic habitat, 
tions in growth are asso- 
ear variation in summer 
ter temperatures. Signifi- 
rowth rates also occur in 
habitats under the same 

The rate of saguaro growth differs widely from 
population to population in response to the broad 
range of climatic environments present within the 
extensive geographic range of the species. 
Generally, rates of saguaro growth decrease along 
a gradient of (1) decreasing summer precipitation 
from east to west, and (2) decreasing winter mini- 
mum temperatures from south to north latitudes, 
and from low to high elevations. 

Freezing causes a significant reduction in 
saguaro growth which can continue for several 
years after such injury. In saguaro populations 
situated near the cold-limited northeastern 
boundary of their distribution, the growth- 
limiting effect of recurring catastrophic freezes 
offsets the advantage of relatively favorable 
moisture conditions. The average growth of large 
saguaros in these environments is less than in 
some more arid but warmer environments. Growth 
rates and, therefore, height-age relationships 
of saguaros growing in different portions of the 
species' range are not only determined by 
differences in precipitation and related avail- 
ability of moisture but are also controlled by 
the frequency and intensity of sub-freezing tem- 
peratures. Established height-age relationships 
for saguaro populations growing in one locality, 
therefore, cannot be applied to populations 
growing in other locations without consideration 
of both of these controlling climatic variables. 

Over a major portion of the saguaro species 
distribution, winter cold determines not only the 
growth rate, but also quantitative aspects of its 
reproductive success, local and absolute limits 
of distribution, population density and age-class 
structure, as well as the timing of the critical 
events in the life cycle of the saguaro. 



Having established the nature and operation of 
the factors that control germination, establish- 
ment, and survival of saguaros in Saguaro National 
Monument and elsewhere in this more northerly por- 
tion of their range, it is possible to draw 
conclusions concerning the current status of these 
populations in relation to past conditions, events, 
and activities, and in some measure to predict 
their future course. 

Historical Factors 

The condition of sagu 
Saguaro National Monumen 
varying degrees to histo 
include removal of young 
chants, woodcutting, gra 
predator control. These 
humans and their domesti 
imposed upon the natural 
that limit the natural d 
saguaro and, within thes 
germination, establishme 

aro populations at 

t has been attributed in 

ric activities. These 

saguaros by cactus mer- 
zing, fire control and 

historical activities of 
c animals have been super- 
environmental factors 
istribution of the 
e limits, control its 
nt, and survival. 

Cactus Pirates 

During the early part of this century, large 
numbers of young saguaros in the vicinity of 
Tucson were removed for sale to cactus fanciers, 
and for landscaping purposes. There is no reli- 
able basis for estimating the number of saguaros 
removed during these operations. That young 
saguaros were present at that time in numbers 
sufficient to support such operators, however, 
leaves no question that conditions were favorable 
for saguaro establishment, and for their survival 
to a size and age suitable for "harvest." It is 
likely that most of the saguaros sought and 
removed were the easily transplantable sizes of 
the 10- to 30-year-old age-group. 

Plants of that age-class have a high natural 
life expectancy. They have grown beyond the age 
of high vulnerability to the principal agents of 
young saguaro mortality, and are entering a long 
period of growth when they are least vulnerable 
to the climatic factors that kill large adult 
plants. Thus, it is reasonable to assume that, 
in the absence of human intervention, a large 
proportion of those plants removed by cactus mer- 
chants would be alive today as intermediate-size 
adult saguaros--the age-group now largely missing 
in the Cactus Forest. 

Woodcut ti ng 

Mesquite (Prosopis juliflora), a preferred 
firewood species, was intensively cut within the 
Cactus Forest area of the monument. The durable 
cut stumps of this species are still abundant 
there, and most of the living mesquite trees are 
sprouted from old stumps. During the latter part 
of the last century the green wood of paloverde 
trees (Cercidium microphy Hum) harvested in the 
Cactus Forest was used to fire limestone kilns 
located within the area. There is, however, no 
record of the number of paloverde trees removed 
for this use. 

saguaros, woodcutting also reduced the number of 
favorable sites for subsequent saguaro establish- 
ment. Without the modifying influence of pro- 
tective plant cover, the associated microhabitats 
were no longer suitable for the establishment of 

Livestock Grazing 

Some portions of Saguaro National Monument have 
been subjected to intensive grazing by cattle for 
more than three-quarters of a century. In some 
localities, particularly the Cactus Forest area 
of the east monument, this has had an unquestion- 
ably detrimental effect on the germination, es- 
tablishment and survival of young saguaros. 

Direct destruction of young saguaros has 
resulted from trampling of cattle seeking shade 
and forage beneath the crowns of desert trees. 
Even more significant has been the impact of con- 
tinued grazing pressure upon the physical struc- 
ture of the plant community. 

The primary role of historic livestock grazing 
has been to reduce the density of affected saguaro 
populations by decreasing the number and quality 
of sites suitable for germination, establishment 
and survival of young saguaros thereby increas- 
ing exposure to natural mortality-causing factors. 
Differential impact, i.e., differential mortality, 
has been associated with differences in the inten- 
sity and duration of grazing and the physical 
characteristics (rockiness) of the terrain. Mor- 
tality from natural factors is highest on inten- 
sively grazed, relatively flat non-rocky terrain, 
and lowest on less intensively grazed steep slopes 
with extensive rock outcrops. Natural regenera- 
tion of vegetation is occurring, and natural rates 
of germination, establishment, and survival of 
young plants are now occurring in habitats under 
National Park Service protection that were heavily 
grazed earlier in the century. 

Fire Control 

Fire is a natural cause of mortality of young 
saguaros near the upper elevational limits of 
their distribution. Our observations on natural 
fire-caused saguaro deaths in the Coronado and 
Tonto National forests in Arizona indicate that 
fire is a density-controlling factor in ungrazed 
habitats where grasses and other low-growing 
plant species provide a sufficiently continuous 
cover of combustible fuel to permit the spread of 
natural fire. 

However, there is no 
caused saguaro deaths w 
Monument. This is not 
policy of controlling a 
directly a consequence 
that naturally would fo 
ous ground cover on the 
been closely cropped, a 
drastically reduced by 
grazing. Fires can nei 
these denuded, grossly 

known instance of fire- 
ithin Saguaro National 
only a result of a strict 
11 fires, but even more 
of cattle grazing. Grasses 
rm a more-or-less continu- 

rocky footslopes have 
nd their density has been 
long-continued cattle 
ther start nor spread in 
altered environments. 

Activities of woodcutters affect the fate of 
young saguaros growing beneath the trees they 
remove. Not only are such saguaros liable to 
mechanical destruction by the woodcutters them- 
selves, the survivors are fully exposed to tram- 
pling by cattle and to the full decimating impact 
of the biotic and abiotic environment. In addi- 
tion to such direct destruction of established 

Predator Control 

Predator control within and adjacent to both 
sections of Saguaro National Monument continued 
into the last decade cannot, with certainty, be 
related to the past and present condition of 
saguaro stands within the boundary. The whole 
question of the effect of predator control on 


fluctuations in rodent populations seems insepa- 
rably related to a host of other factors ultimately 
of equal or greater importance in the dynamics of 
desert rodent populations. These include food and 
water supply, habitat modifications produced by 
livestock grazing, climatic cycles, and the result- 
ing complex interactions including predation. 

Ultimately, it seems that the most that can be 
said with regard to predator control is that to 
some degree it may have allowed populations of 
some prey species to increase until some other 
environmental factor became limiting on their num- 
bers. There are no data that show that an increase 
in rodent predators on young saguaros actually 
resulted from predator control, that, rodent popu- 
lations were maintained at unnaturally high levels, 
or that the establishment of young saguaros was 
unnaturally limited by high populations of rodents 
resulting from the control of predators. Whatever 
the previous consequences of predator control on 
these saguaro populations, the present condition 
within the two sections of the monument appears 
to be that of an essentially natural relationship. 

Responses of an Abused Environment 

The role of histo 
act as new and overr 
torical factors have 
gate the operation o 
Acting primarily in 
all effect has been 
suitability of sites 
ment and survival of 
Saguaro National Mon 
been eliminated from 
by these activities, 
the density of large 
plants is below the 
habitats . 

rical factors has been not to 

iding controls; rather, his- 
acted to intensify or miti- 

f natural regulating factors. 

an indirect manner, the over- 

to decrease the number and 
for germination, establish- 
young saguaros. Within 

ument, the saguaro has not 
any of its former habitats 
but rather, in some habitats, 
juvenile and young adult 

expected level for these 

There is no cause to believe that historical 
factors have created irreversible deterioration 
of any of the primary habitats of the saguaro at 
Saguaro National Monument. Where elimination of 
adverse uses has been accomplished it has been 
followed by natural regeneration of the plant com- 
munity, and by the establishment of large numbers 
of juvenile saguaros. 

In contrast to the condition reported only a 
few years ago, and still presumed by many tq be 
the situation, there are today in each of the two 
sections of Saguaro National Monument thousands of 
young saguaros less than a half -meter (20 inches) 
tall. These occur not only on the rocky foot- 
slopes of the Rincon and Tucson mountains—where 
the presence of substantial numbers of large 
juvenile and young adult saguaros indicates con- 
tinued establishment and survival of young 
saguaros in these habitats — but notably in the 
non-rocky flat habitats of the Cactus Forest area, 
and the lower bajadas of the respective east and 
west sections of the monument. 

Hypotheses of Population Change 

Some saguaro populations fluctuate markedly in 
density and age-structure. This is especially 
true and even characteristically so in the cold- 
limiting parts of the range of this subtropical 
species (see Harper 1967; Steenbergh and Lowe 
1976a). Other investigators have offered, in 
various combinations, three basically differing 
hypotheses to explain observed fluctuation in 
saguaro populations in various localities includ- 
ing Saguaro National Monument. There are (1) 
over-grazing, (2) climatic change, and 

(3) "disease." However, as noted below, none 
of these hypotheses adequately accounts for the 
operation of the critically important climatic 
factors that ultimately limit the distribution 
and control the age-structure of those popula- 
tions that are situated at the limits of the 
species' distribution. 

The grazing-plus-rodents hypothesis suggests 
that in some habitats high rodent populations 
resulting from grazing-caused alterations of the 
plant community destroy young saguaros, causing 
a slow decline and disappearance of saguaro popu- 
lations in these habitats. Obviously, over- 
grazing produces changes in the natural environ- 
ment that are seriously detrimental to the 
establishment and survival of young saguaros. 
However, the hypothesis may overestimate the ulti- 
mate impact of rodents, and fails to accord 
sufficient importance to the controlling effect 
of recurring catastrophic freezes in these 
marginal habitats. 

Climatic change, 
higher temperatures 
with grazing) has a 
of the lack of youn 
portions of Saguaro 
the observed change 
along the eastern a 
distribution- -where 
by lack of moisture 
be explained either 
or summer precipita 
winter temperatures 

a shift toward slightly 
and lower rainfall (together 
lso been proposed as a cause 
g saguaros in the "valley" 

National Monument. However, 
s in the saguaro populations 
nd northern boundaries of its 
the species is limited neither 
nor high temperature--cannot 
by a decrease in mean winter 
tion, or by a rise in mean 

"Bacterial necrosis disease" has also been 
proposed as a primary cause of saguaro population 
decline at Saguaro National Monument. The so- 
called "disease," however is a result — not a 
cause-- of saguaro death. Bacterial rot is the 
long-understood subsequent and ecologically im- 
portant process of natural decomposition of soft 
tissues of saguaros that are killed outright or 
rendered physiologically dysfunctional by mechani- 
cal factors, primarily freezing and lightning 
(see Lowe 1964, 1966; Steenbergh 1972; Steenbergh 
and Lowe 1976a, b) . Reported "epidemics of bac- 
terial necrosis disease of the saguaro" are the 
inevitable response of the populations to recur- 
ring catastrophic freezes that are a normal 
characteristic of the regional climate. 

Limiting Factors 

The saguaro giant cactus is not some rare 
exception to the fact that environmental limiting 
factors modify all terrestrial plant populations 
wherever located and, at the ultimate limits 
of the species distribution, such factors finally 
become 100% limiting. Our investigations and 
the work of others leave no doubt that natural 
environmental factors are the underlying cause 
of the dramatic changes that have occurred within 
the saguaro populations at Saguaro National 
Monument and elsewhere in the northern portion 
of its range. 

Saguaro Nationa 
just near the sagu 
the saguaro specie 
the monument. The 
a marginal one in 
particular populat 
which occupy margi 
topographic situat 
tures are intensif 
accumulation . 

1 Monument (east) is not only 
aro's distribution limits, 
s boundary line runs through 
east monument population is 
every sense of the term. The 
ions in question are those 
nal saguaro habitats — 
ions where subfreezing tempera- 
ied by cold air drainage and 


Among the many physical and biotic factors 
acting upon these populations, it is climate — 
namely the extremes of winter cold — that is 
clearly the overriding factor in the regulation 
of these populations. 

Populations in Response to 
Climatic Events 

Diverse records on saguaro populations (and 
other warm-desert plant populations) and on cli- 
matic events clearly tell the story of populations 
in trouble since the middle of the last century. 

The presence o 
near the base of 
during the latter 
indicated by the 
and Hubbard (1899 
commented general 
young saguaros 
numerical data on 
and concluded tha 
maintaining thems 

f dense popul 
the mountains 

part of the 
) . MacDougal 
ly on the dif 
Shreve (1910) 

saguaro popu 
t the populat 
elves . 

ations of saguaros 
east of Tucson 

19th century are 

of Bigelow (1887) 
(1908) , however, 

ficulty of finding 
offered the first 

lacions at Tucson 

ions were not 

The saguaro population in portions of Saguaro 
National Monument (east) was in a long-standing 
state of decadence at the time of the establish- 
ment of the monument in 1933. As seen in the 
photographs by Homer Shantz 5 (also see Hastings 
and Turner 1965) there is an abundance of large, 
old saguaros, and a conspicuous sparsity of 
juvenile and young adult saguaros. The very 
characteristic that inspired Shantz (1932) to say 
that "Nowhere in the world is there so fine a 
stand of giant sahuara (Carnegia gigantia) as in 
the University Cactus Forest" was the irrevocable 
condition that was to lead to the subsequent and 
inevitable dramatic decline of that same stand. 

Analysis of sagua 
National Monument ca 
Bureau of Plant Indu 
1946; Gill 1951; Mie 
May 1962; Steenbergh 
earlier observations 
and Wilder (1940) th 
in the east monument 
insufficient to main 
highest densities of 
rocky footslopes. 

ro populations at Saguaro 
rried out in 1941 by the USDA 
stry (Gill and Lightle 1942, 
Ike 1944; also see Alcorn and 

and Lowe 1976a) support 

of Wilder and Wilder (1939) 
at the number of young saguaros 

flats stand had long been 
tain the stand, and that the 

saguaros occurred on the 

Investigations in 1962 and 1963 
Whittaker and Lowe 1963; Lowe 1964 
Whittaker 1965) on saguaro populat 
vicinity of Tucson further confirm 
vations that saguaro establishment 
rates in rocky habitats are higher 
finer soil of non-rocky habitats ( 
1962; Bingham 1963). More importa 
studies further support the earlie 
by Thornber (1916) that massive di 
is associated with catastrophic fr 
plain the relationship of freezing 
higher survival rates in rocky hab 
data in those reports, it can also 
f reeze-caused saguaro mortality de 
east to west along a gradient of i 
warmer winter climate. 

(Niering , 

; Niering ana 

ions in the 
earlier obser- 
and survival 
than in the 

also see Kramer 

nt, these 

r observations 

e-off of saguaros 

eezing, and ex- 
weather to 

itats. From the 
be seen that 

creases from 


Further data on saguaro populations in Saguaro 
National Monument and the relationship of saguaro 
size and habitat to f reeze-caused mortality, are 
reported by Steenbergh and Lowe (1976a). In that 

Homer L. Shantz photo collection, University 
of Arizona Herbarium. 

report, we conclude that catastrophic freezing 
selectively structures saguaro populations remov- 
ing the smallest (youngest) and the largest 
(oldest) plants, leaving a high percentage of 
large juvenile and unbranched young adult 
saguaros with heights from 0.46 to 3.80 m 
(1.5-12.5 ft) . 

Catastrophic Freezes 

Daily weather observations provide a record of 
minimum temperatures in Tucson since 1895 (see 
Steenbergh and Lowe 1976a, Fig. 25). Other 
diverse records of these climatic events, and the 
response of saguaro populations and other cold 
sensitive plant populations are scattered through- 
out the literature (Thornber 1912, 1916; Wiggins 
1937; Turnage and Hinckley 1938; Niering et al. 
1963) , and are, in fact, recorded in the respon- 
sive plant populations themselves. 

Saguaro populations in the Tucson vicinity are 
an unexamined record of catastrophic freeze 
occurrence that long predates the oldest written 
weather records. The unbalanced size-class struc- 
ture of these and other saguaro populations is an 
indication of past winter climate. In many 
northern saguaro populations constrictions near 
the base of the stem tell the story of previous 
catastrophic freezes. These constrictions re- 
sult from f reeze-caused crown-kill. In some 
habitats at Saguaro National Monument such scars, 
present on nearly every large juvenile and adult 
saguaro, provide a continuous record of recurring 
catastrophic freezes as old as the plants them- 
selves. Based on the estimated age of larger 
saguaros in these habitats, recurring catastrophic 
freezes have been a normal part of these environ- 
ments for more than 150 years. 

Low temperature readings in Tucson recorded 
during the first decade of this century were 
followed by MacDougal's (1908) observation on 
the "...abundance of fallen skeletons...." The 
most noteworthy record of low temperature is the 
6°F (-14.4°C) Tucson temperature recorded on 
January 13, 1913. As the first report on cata- 
strophic freeze-kill of saguaros, Thornber' s 
(1916) observations on the devastating effects of 
that freeze are particularly significant: 

". . . during the very cold winter of 1912- 
13 thousands of small giant cactus plants 
growing near their greatest altitudinal 
limits were killed outright. Many other 
species also suffered great damage." 

The catastrophic freeze that occurred in January 
1937 , and the resulting severe damage to warm-desert 
plants, are the subject of a comprehensive re- 
port by Turnage and Hinckley (1938) . Although 
damage to saguaros was not noted in that report, 
subsequent events leave no doubt that it was 
among the species most severely damaged by that 
freeze. The spectacular die-off of saguaros 
first noted in 1939 and continuing into the 1940' s 
(Gill and Lightle 1942, 1946; Mielke 1944; Gill 
1951) was the delayed response to the subfreezing 
temperatures of January 1937, the coldest period 
in 24 years (Turnage and Hinckley 1938) . 

It is well known that not all parts of a plant 
die at the same moment. This time scale of organ 
death is greatly exaggerated in the saguaro giant 
cactus. As many as nine years may elapse between 
lethal injury, and collapse of a freeze-damaged 
saquaro. Thus, the death of large saguaros 
lethally injured in 1937 escaped recognition until 
1939 and later, when with their decay and collapse 


death became obvious. 

Saguaro die-off in response to catastrophic 
freeze followed by rapid thawing conditions in 
1962 reported by Niering et al. (1963) and Lowe 
(1964) was followed by rapid collapse of many 
of the stricken individuals (see de Candolle 
1852). However, as in the case of the 1937 
freeze, the death of many plants did not become 
evident for one or more years after the freeze. 
The 1962 freeze also damaged saguaros and other 
plants at Organ Pipe Cactus National Monument 
where we observed resulting severe injury to 
senita (Cereus schotti) and organpipe cacti 
(Cereus thurberi) , as well as to other cacti, 
elephant trees (Bursera microphylla) and other 
subtropical shrubs. 

The initial response of saguaro populations to 
the freeze of January 1971 is described by 
Steenbergh and Lowe (1976a) and resulting col- 
lapse of saguaros has continued into 1976. We 
also observed, as a result of the 1971 freeze, 
damage to organpipe and senita cacti at Organ 
Pipe Cactus National Monument, and near Tucson, 
the death of other species of cacti, namely 
bisnaga [Ferocactus wislizeni) and the chain- 
fruit cholla (Opuntia fulgida) . In the Tucson 
vicinity, we also observed severe damage to 
desert ironwood {Olneya tesota) , and foothill 
paloverde (Cercidium microphyllum) that in some 
localities resulted in the subsequent death of 
numerous individuals of both species. 

It is evident from these diverse records that 
frequently recurring extremes of subfreezing tem- 
peratures are a long-standing normal character- 
istic of the climatic environment of the northern 
portions of the Sonoran Desert. Further, it is 
evident that these events exert a powerful in- 
fluence not only on populations of the saguaro, 
but on populations of other cold-limited Sonoran 
Desert plant species as well. Acting in a cata- 
strophic manner, such freezes are the primary 
control on the limits, local distribution and 
dynamics of saguaro populations at Saguaro 
National Monument and elsewhere in the colder 
portion of the species' distribution in Sonora 
and Arizona. Both the dramatic die-offs of large, 
old saguaros and the lack of younger replacements 
observed in these populations are the direct and 
inevitable response of these populations to the 
repeated occurrence of these devastating climatic 


The outlook for the future of saguaros at 
Saguaro National Monument is not an optimistic 
one. Neither, however, does it conform to the 
grim predictions of early extinction that some 
have offered. Barring unlikely sudden environ- 
mental changes in the habitats of its major 
occurrence within the monument, the saguaro will 
continue to survive long after our concern with 
the question is terminated. ' 

There are today thousands of young saguaros 
at Saguaro National Monument, many of which will 
survive beyond a human lifetime. Many more of 
these, however, will perish during that time. 
Certainly, there will not be in our time--and 
perhaps never again — a return to the condition 
that inspired Homer Shantz (1932) to describe it 
as the finest stand of the giant saguaro in the 

Generally, even in the absence of further 
adverse human influences, we can expect that the 

density of large saguaros in most, if not all, 
habitats within the monument will be lower than 
in the recent past. In some habitats, however, 
we can expect that with the complete elimination 
of livestock grazing an increase in the density 
of young saguaros will accompany a natural regen- 
eration of the damaged plant communities. 

Within the monument, continuing short term 
climatic variation can be expected to produce 
corresponding fluctuations of saguaro populations — 
but not to cause their early extirpation. 
Generally, this will effect changes in density, 
with unfavorable winter temperatures eliminating 
young plants from the least favorable microsites. 

Saguaro populations, and other cold-limited 
Sonoran Desert plant species will continue to be 
controlled in the colder portions of their dis- 
tributions by recurring catastrophic freezes. 
Any change in the frequency, intensity or duration 
of these recurring catastrophic freezes will re- 
sult in a response by the affected populations. 
Any trend in these populations will follow the 
trend of those particular parameters of the win- 
ter climate that exert the greatest effect on the 
youngest members of the plant populations. Our 
lack of knowledge of long-term climatic trends 
and our inability to predict the occurrence of 
such critical climatic events, however, precludes 
long-range projections concerning the ultimate 
fate of these populations. 

With its high reproductive potential and long 
life-span, the saguaro is well adapted to maintain 
itself in an environment in which occasional 
freezes that result in catastrophic die-off are 
followed by regeneration of the population during 
intervening periods of climatic remission. We 
have no information indicating that such climatic 
events have not similarly affected many past gen- 
erations of saguaros in the same habitats where 
they presently occur. It is entirely possible 
that we have observed in these northern saguaro 
populations but one phase of the normal fluctua- 
tion of populations whose stability must be 
measured not in years, but in generations. 

It may well be that 
observed — the "decline 
lations--is but a limi 
span too brief to perm 
adaptive strategy of a 
by natural selection h 
system for survival in 
catastrophic climatic 
planation is correct, 
neither in biology nor 
limited perspective, a 
in the perspective off 

the "problem" that we have 
" of specific saguaro popu- 
tation imposed by our life- 
it our recognition of an 

longer lived species that 
as evolved a near-perfect 

an environment of recurring 
events. If that likely ex- 
then the real "problem" is 

management, but in a 
nd the only "solution" lies 
ered by the time scale of 


ALCORN, S. M. , and C 
saguaro forest. Pi 

BENSON, L. 1969. The 
Univ. Arizona Press 

BIGELOW, J., JR. 1887 
ment XIII. Outing 

BINGHAM, S. B. 1963. 
ships in two stands 
community of the So 
Univ. of Arizona, Tu 

vegetables. Am. J. 
Intell. #7). 

MAY. 1962. Attrition of a 
ant Dis. Rep. 46 (3) : 156-158 . 

cacti of Arizona. 3rd ed . 
, Tucson. 218 pp. 

After Geronimo. Install- 
9(6) :522. 

Vegetation-soil relation- 

of the Cere idium-Caznegiea 
noran Desert. M.S. Thesis, 
cson. 84 pp. 
852. On the freezing of 

Sci. Arts II 14:445(Misc. 


GILL, L. S. 1951. Mortality of the giant cactus 
at Saguaro National Monument, 1941-1950. Sa- 
guaro National Monument Headquarters (Tucson, 
Arizona), Official Rep. 5 pp.; 2 tables; 
1 fig. 

GILL, L. S., and P. C. LIGHTLE. 1942. Cactus 
disease investigations. Saguaro National Mon- 
ument Headquarters (Tucson, Arizona) , Official 
Rep. 40 pp.; 9 tables; 15 figs. 

GILL, L. S., and P. C. LIGHTLE. 1946. Analysis 
of mortality in saguaro cactus. Saguaro Na- 
tional Monument Headquarters (Tucson, Arizona) , 
Official Rep. 4 pp.; 11 tables. 

HARPER, J. L. 1967. A Darwinian approach to 

plant ecology. J. Anim. Ecol. 36 (3) : 495-518 . 

HASTINGS, J. R., and R. M. TURNER. 1965. The 
changing mile: An ecological study of vege- 
tation change with time in the lower mile of an 
arid and semiarid region. Univ. Arizona Press, 
Tucson. 317 pp. 

1972. Carnegiea gigantea . Page 63 in J. R. 
Hastings, R. M. Turner, and D. K. Warren, An 
atlas of some plant distributions in the So- 
noran Desert. Univ. Ariz. Inst. Atmosph. 
Physics Tech. Rep. Meteorol. Climatol. Arid 
Regions No. 21. 

HUBBARD, H. G. 1899. Insect fauna of the giant 
cactus of Arizona: Letters from the Southwest. 
Psyche 8(Suppl. 1):1-14. 

KRAMER, R. J. 1962. The distribution of saguaro 
(Cereus giganteus Engelm.) in relation to 
certain soil characteristics. M.S. Thesis. 
Ariz. State Univ., Tempe . 120 pp. 

LOWE, C. H. 1964. Arizona landscapes and habi- 
tats. Pages 1-132 in Charles H. Lowe (ed.), 
The vertebrates of Arizona. Univ. Arizona 
Press, Tucson. 

LOWE, C. H. 1966. Life and death of the sahuaro 
in Arizona. Cactus Capital Chatter 1(8): 2-3. 

MACDOUGAL, D. T. 1908. Across Papagueria. Plant 
World 11(5) :93-99; 11 (6) : 123-131 ; (Also Am. 
Geogr. Soc. Bull. 40:1-21.) 

MIELKE, J. L. 1944. Summary of results of con- 
trol experiments on saguaro disease, Saguaro 
National Monument. Saguaro National Monument 
Headquarters (Tucson, Arizona), Official Rep. 
4 pp. 

NIERING, W. A., and R. H. WHITTAKER. 1965. The 
saguaro problem and grazing in southwestern 
National Monuments. Natl. Parks Mag. 39(213): 
4-9; illus. 

1963. The saguaro: A population in relation 
to environment. Science 142 (3588) : 15-23 . 

SHANTZ, H. L. 1932. Description. in Roger W. 
Toll, Untitled report to the Director of Na- 
tional Park Service. 

SHELTON, N. 1972. Saguaro National Monument, 
Arizona. U. S. Dept. Int., Natl. Park Serv. 
Nat. Hist. Ser. 98 pp. 

SHREVE, F. 1910. The rate of establishment of 
the giant cactus. Plant World 13 (10) ; 235-240 . 

SHREVE, F. 1931. Fouquieriaceae , Larrea triden- 
tata, Carnegiea gigantea . Die Pf lanzenareale , 
Ser. 3 (1) :4-6; 3 maps. 

SHREVE, F. 1951. Vegetation of the Sonoran 
Desert. Carnegie Inst. Wash. Publ. 591. 
192 pp.; maps and photos. 

SHREVE, F., and I. L. WIGGINS. 1964. Vegetation 
and flora of the Sonoran Desert. Stanford 
Univ. Press, Stanford, Calif. 2 vols. 1740 pp. 

STEENBERGH, W. F. 1970. Rejection of bacterial 
rot by adult saguaro cacti (Cereus giganteus ) . 
J. Ariz. Acad. Sci. 6(1):78-91. 

STEENBERGH, W. F. 1972. Lightning-caused de- 
struction in a desert plant community. South- 
west. Nat. 16(3/4) :419-429. 

STEENBERGH, W. F., and C. H. LOWE. 1969. Criti- 
cal factors during the first years of life of 
the saguaro (Cereus giganteus ) at Saguaro 
National Monument, Arizona. Ecology 50(5) :825- 

STEENBERGH, W. F., and C. H. LOWE. 1976a. 
Ecology of the saguaro. I. The role of 
freezing weather on a warm-desert plant popu- 
lation. Pages 49-92 in Research in the Parks. 
National Park Service symposium series no. 1. 
Government Printing Office, Washington, D.C. 
In Press. 

STEENBERGH, W. F., and C. H. LOWE. 197 6b. 

Ecology of the saguaro. II. Reproduction, 
germination, establishment, growth, and sur- 
vival of the young plant. U.S. Dept. Int., 
Natl. Park Serv. Scientific Monogr . Series 
No. 8. In Press. 

THORNBER, J. J. 1912. Resistance to frost of 
introduced trees and shrubs. Ariz. Agric. 
Exp. Stn. Timely Hints for Farmers No. 91. 

THORNBER, J. J. 1916. Introduction. Pages 

119-122 in J. C. T. Uphof, Cold-resistance in 
spineless cacti. Univ. Ariz. Agric. Exp. Bull. 

TURNAGE, W. V., and A. L. HINCKLEY. 1938. 

Freezing weather in relation to plant distri- 
bution in the Sonoran Desert. Ecol. Monogr. 

WIGGINS, I. L. 1937. Effects of the January 
freeze upon the pitahaya in Arizona. Cactus 
Succulent J. 8:171. 

WILDER, C. S., and J. C. WILDER. 1939. Re- 
establishment of saguaros. Southwest. Mon. 
Spec. Rep. 26:153-160. 

WILDER, J. C. 1940. Saguaros, old and young. 
Desert Plant Life 12(4):65-66. 


Ronald I. Miller and Larry D. Harris 

A relation between the size of an area and the 
total number of species in any taxon was documented 
for both plant and animal communities at the be- 
ginning of this century. Plant ecologists were 
the first to notice that a relative increase in 
species number correlates well with an increase 
in quadrat size (Arrhenius 1921, Gleason 1922) . 
A concurrent study noted a similar relation be- 
tween faunal species number and sample size 
(Willis 1922) . This relation was later formalized 
with the introduction of the species-area curve 
(figure 1) . This allometric formulation demon- 
strates that the incremental increase in the num- 
ber of species recorded in an area declines for 
larger areas. In other words, an increase in a 
larger area involves a smaller species accrual 
than a similar increase in a smaller area. 

relation be 
and the siz 
among the f 
observed th 
on islands 
every 10-fo 
This empiri 
fied by Pre 
and Wilson 

nt observati 
tween the ob 
e of an isla 
irst to arti 
at the numbe 
in the West 
Id increase 
cal relation 
ston's (1962 
and thereaf 
1967) as: 

S = CA 5 

ons uncovered a unique 
served number of species 
nd . Darlington (1957) was 
culate this relation. He 
r of reptile species found 
Indies increased by 50% for 
in island (figure 2) size, 
was independently veri- 
) model of finite species 
ter generalized (MacArthur 


where S is the predicted number of species; C is 
a parameter primarily dependent upon the species 
diversity of the taxon being considered, the zoo- 
geographic region and the underlying environmental 
heterogeneity; A is the area of the island and z 
represents the degree of relation between the 
species number and the size of the area. The 
validity of this equation for quantifying species 
numbers on both oceanic and continental islands 
has been empirically verified (Preston 1962, 
MacArthur and Wilson 1967, Vuilleumier 1970, Brown 
1971, Terborgh 1974, 1975; Diamond 1975, Moore and 
Hooper 19 75) . The predictive power of this 
equation is based on both the geographic and eco- 
logical characteristics of an area. 

The distinction between insular and sample 
communities is demonstrated by the empirical 
observation that isolated areas exhibit z-values 
[from equation (1)] between 0.20-0.35 while areas 
representing only samples of larger communities 
exhibit z-values ranging from 0.11-0.17 (Preston 
1962) . Differences in the number of species in- 
habiting island areas versus continental areas 
are explicable from this distinction. 

The species contai 
areas contiguous with 
sent a sample from a 
These groups of speci 
by their interactions 
thus they do not nece 
contained communities 
communities of insula 
characterized by an i 

ned by demarcated continental 
surrounding habitat repre- 

larger community (figure 3) . 

es will be greatly influenced 
with surroundi ig species and 

ssarily represent self- 

On the other hand, species 

r areas represent communities 

ndependence from contiguous 

School of Forest Resources and Conservation, 
University of Florida, Gainesville. 

environments. Thus insular communities have pre- 
sumably developed an internal integrity lacking 
in sample areas. For this reason a greater 
species-area relation is observed in insular com- 
munities. Moreover, fewer species are found in 
insular areas than in sample areas of equal size, 
and therefore species communities on islands are 
comparatively depauperate. 

It has recently been proposed that a species- 
area relation similar to the one found on islands 
may occur in continental areas separated from 
their surrounding gene pools (Diamond 1972, 1975; 
Hooper 1971, May 1975; Sullivan and Shaffer 1975, 
Terborgh 1974, 1975, Wilson and Willis 1975) . 
This hypothesis is supported by a number of inde- 
pendent lines of observation. First, the number of 
species in naturally occurring isolated areas such 
as mountaintops (Brown 1971) and alpine grasslands 
(Vuilleumier 1970) is predicted by an insular 
species-area relation. Secondly, an island model 
explains a significant portion of the extinction 
rate recorded in the recently isolated avifaunal 
community of Barro Colorado Island (Willis 1974, 
Terborgh 1975) . Finally, forest size is a signi- 
ficant predictor of the number of bird species in 
recently created "forest islands" in developed 
areas in both the United States (Galli et al. 
1976) and in Great Britain (Moore and Hooper 1975). 

As wildlife preserves and national parks be- 
come surrounded by "seas of civilization" they 
may begin to reach a condition similar to isolated 
continental areas (figure 4) . As these protected 
areas transform from sample areas to islands one 
would expect the animal communities to change 
accordingly. Since sample continental areas 
generally contain more species per unit area than 
islands, newly demarcated preserves might be con- 
sidered to be "supersaturated" with species. 
This implies that the transformation process is 
tantamount to species extirpation. As the pre- 
serve becomes progressively more isolated from the 
surrounding natural habitat the species ensemble 
will slowly be reduced and eventually achieve a 
lower number of species appropriate to the more 
insular condition. 

The Theory of Island Biogeography asserts that 
the size and characteristics of island communities 
are determined by a dynamic equilibrium between 
species immigration and extinction rates (figure 
5) . The theory predicts that the variable pro- 
cesses of immigration and extinction on islands 
will produce a constant turnover of species. 
The relative effect of these processes will be 
based on the distance of the island from the 
nearest species pool and island size. This 
dynamic equilibrium will produce a characteristic 
number of resident species for each island. As 
yet, this model has not categorically been shown 
to occur on either island or isolated continental 
areas (Simberloff 1976) . 

We believe the applicability of the island con- 
cept to areas such as national parks and pre- 
serves, fast becoming isolates because of en- 
croaching civilization, is solely predicated on 
the empirically based species-area phenomenon. 





o I 6- 




CD I 4- 



I 2H 


FIGURE 1. The number of recorded species (ordinate) 
plotted against areas of increasing size (abcissa) 
produces this characteristic allometric function 
known as the species area curve. 

4 5 6 

Area (A) 




FIGURE 2. For islands, Darlington (1957) 
observed that a tenfold increase in area 
represents a 50% increase in the number 
of reptile species. 

90% Reduction in Area 

507o Reduction in Species 

Isolated Area 


FIGURE 3. A demonstration of the distinc- 
tion between insular and sample communities. 
Insular communities are self-contained, 
whereas sample communities interact with 
surrounding species. 

FIGURE 4. National parks and 
preserves are becoming isolated 
from surrounding natural area 
by encroaching civilization. 



FIGURE 5. Equilibrium model for the biota of a 
single island (MacArthur and Wilson 1967) . 
The equilibrium number of species, S, is deter- 
mined by a dynamic relation between immigration 
and extinction rates. 

Number of Species Present, N 

Thus it is not necessary to invoke the unverified 
Theory of Island Biogeography to predict that a 
depauperization will probably occur in isolated 
parks and preserves. The dynamic equilibrium 
model only proposes a mechanism to explain the 
species-area phenomenon on islands. 

Verification of the depauperization process in 
isolated preserves, predicated on the species- 
area relation, will most likely result from 
rigorous surveys of the species ensembles and 
populations within specific areas. In addition, 
research directed toward understanding the ecology 
in isolated communities will lead to predictions 
of species specific vulnerability to extirpation 
pressures within isolated preserves . Data for 
testing recent hypotheses (Diamond 1975, Wilson 
and Willis 1975) concerning the design and manage- 
ment of isolated communities are badly needed. 
Most importantly, understanding the possible ef- 
fects of insularity in national parks and preserves 
will aid conservationists in their attempt to 
mitigate the extinction process in these areas. 


GLEASON, H. A. 1922. On the relation between 
species and area. Ecology 3:158-162. 

HOOPER, M. D. 1971. The size and surroundings of 
nature reserves, pp. 555-561. in The scien- 
tific management of animal and plant communi- 
ties for conservation. E. Duffy and A. S. 
Watt (eds.). Blackwell, Oxford, England. 

MACARTHUR, R. H. and E. 0. WILSON. 1967. The 
theory of island biogeography, Princeton Uni- 
versity Press, Princeton, 203 p. 

MAY, R. M. 1975. Patterns of species abundance 
and diversity, pp. 81-120, in Ecology and evo- 
lution of communities. M. L. Cody and J. M. 
Diamond (eds.). Harvard Univ. Press, Cam- 

HOOPER. 1975. On the num- 
in British woods. Biol. 

The canonical distribution 
Ecology 43(2), p. 


Species and area. J. Ecol. 


BROWN, J. H. 1971. Mammals on mountaintops : non- 
equilibrium insular biogeography. Amer. Natur. 
105(945) :467-478. 

DARLINGTON, P. 1957. Zoogeography. John Wiley 
and Sons, New York, 675 p. 

DIAMOND, J. 1972. Biogeographic kinetics: Esti- 
mation of relaxation times for avifauna of 
Southwest Pacific islands. Proc. Nat. Acad. 
Sci. U.S.A. 69(11) :3199-3203. 

DIAMOND, J. 1975. The island dilemna: Lessons 
of modern geographical studies for the design 
of nature preserves. Bio. Conserv. 7:129-146. 

GALLI, A. E., C. F. LECK, and R. T. FORMAN. 1976. 
Avian distribution patterns in forest islands 
of different sizes in central New Jersey. 
Auk 93:356-364. 


MOORE, N. W. and M. D. 
ber of bird species 
Conserv. 8:239-250. 

PRESTON, F. W. 1962. 

of commonness and rarity 
185-215, and p. 410-432. 

SIMBERLOFF, D. S. 1976. Species turnover and 
equilibrium island biogeography. Science 194: 

SULLIVAN, A. F. and M. L. SHAFFER. 1975. Bio- 
geography of the megazoo Science. 189:13-17. 

TERBORGH, J. 1974. Preservation of natural 
diversity: The problem of extinction prone 
species. Bioscience 24 (12) : 715-722. 

VUILLEUMIER, F. 1970. Insular biogeography in 
continental regions. I. The Northern Andes of 
South America. Amer. Nat. 104 (938) : 378-388 . 

WILLIS, E. O. 1974. Populations and local ex- 
tinctions of birds on Barro Colorado Island, 
Panama. Ecol. Monogr . 44:153-169. 

WILLIS, J. C. 1922. Age and area. Cambridge 
Univ. Press, Cambridge, 259 p. 

WILSON, E. O. and E. 0. WILLIS. 1975. Applied 
biogeography, pp. 522-536. in Ecology and 
evolution of communities. M. L. Cody and J. 
Diamond (eds.). Harvard Univ. Press. Cam- 



A. R. Weisbrod 

The broad stretches of life forms on the North 
American continent may appear to the casual observer 
as a monotonous superficially unbroken expanse of 
countryside. A more careful observer notes that 
each of the great biomes is in fact a complex mo- 
saic of discrete areas, each separated from simi- 
lar units by distinctly different habitats. The 
encroachment of modern civilization has reduced the 
original patterns to broken patches, isolated from 
one another by urban and agrarian activities and 
their connecting transportation routes. Some rem- 
nants of the original habitats have been set aside 
as parks, forests, refuges, and other similar 
natural areas held in common for public use by fed- 
eral, state, and local governments, and by private 
individuals and organizations. Such preserved lands 
are effectively surrounded by a sea of civilization. 
The establishment, management, and use of such re- 
serves must be accomplished with a firm knowledge 
and understanding of the ecological consequences 
of such isolation, if their biotas are to remain 

In the now famous book, The Theory of Island 
Biogeography, MacArthur and Wilson (1967) claimed 
that most areas are isolated to a greater or les- 
ser extent from similar discrete areas. This prop- 
erty of insularity has various measurable influ- 
ences on the structure of biotic communities within 
a given region. These influences include empirical 
relationships such as: the number of species pres- 
ent in the area is directly proportional to its 
size, the farther a given site is from a propa- 
gule source the fewer the species that may either 
reach or survive in the area, and the number of 
species a given area ultimately supports represents 
a dynamic equilibrium between immigration and ex- 
tinction rates. These main points of their notions 
have been greatly reviewed, elaborated, and refined 
by a host of authors. For example, Gilroy (1975) 
suggested that immigration rates are affected pro- 
portionately by species not present in the area while 
extinction rates are proportional only to the spe- 
cies in the area. Diamond (1973) describes a phen- 
omenon he terms "relaxation" in which previously 
settled areas which have become relatively isolated 
begin to lose marginal species as the equilibrium 
number adjusts downward. This adjustment is a re- 
sponse to either falling migration, rising extinc- 
tions, or both. As Wilson and Willis (1975) note, 
such relaxation first affects the large species or 
the specialized forms which by their nature often 
occur in low densities, and they require relatively 
large areas to survive. The effects of this trun- 
cation on these trophic relations concomitant with 
extreme isolation can result in the domination of 
relatively small areas by only a few generally op- 
portunistic forms such as rats (Rattus spp.), rab- 
bits (Oryctolagus cuniculus) , and starlings (stur- 
nus spp.) . These phenomena lead one to consider 
the possibility that national park areas may even- 
tually come to act as ecological islands and some 
may already be experiencing relaxation. 

We can now briefly examine two national park 
examples which may be demonstrating the results of 
the above influences: San Juan Island National 
Historical Park is located on San Juan Island in 
Washington State. The Island is the second largest 

National Park Service, Cooperative Park Studies 
Unit. University of Washington, Seattle. 

island in the San Juan archipelago that separates 
Puget Sound from the Georgia Straights. The is- 
lands form a geomorphic arch across the northern 
end of the Puget Lowlands. Three islands: Orcas, 
San Juan, and Lopez account for 80% of the archi- 
pelago's landmass while the remaining mass con- 
sists of about 169 smaller named islands. The 
archipelago was covered by the great cordilleran 
ice sheets that flowed from the mountains of 
western British Columbia into the Puget Lowlands 
during the Pleistocene. At its greatest extent 
some 15,000 years BP, the glaciers buried the is- 
lands under 1800 meters of ice. 

Recent studies of the islands' mammal fauna, in- 
formation gleaned from scanty literature, and 
our own studies show that San Juan Island contains 
approximately 19 species of land mammals while the 
archipelago contains a total of 24 species (Table 
1) . Bounded within the two units of San Juan Is- 
land Historical Park (SAJH) may be found 14 mammal 
species. The notion that species number is pro- 
portional to the area is immediately evident if one 
plots the number of species against the logarithm 
of the geographic area, i.e., the species-area 
curve (Figure 1) . Note that the Park (SAJH) fits 
comfortably on the curve. The actual number of 
species on the island and probably the Park has 
changed over time and such turnovers in species 
composition are consistent with the stated biogeo- 
graphic influences (Table 2) . The form of the 
changes, that the large and/or specialized species 
are those lost from the fauna, while the replacing 
species are the small generalized colonizers, seems 
to follow from the expectations, derived from known 
impacts of insularity, thus, giving the impression 
that the Park is functioning as an island. Alter- 
natively, one could surmise that the fit of the 
Park to the species-area curve and the turnover 
data are simply artifacts of sampling and bear no 
theoretical relevance. That is, any comparable 
700-hectare area on San Juan Island may show simi- 
lar results. Unfortunately, there is no comparable 
area nor did the factors of time and money permit 
searching for such lands elsewhere in the archi- 
pelago. The resolution of this problem must await 
studies in future decades. 

The second national park presents a different 
situation in several respects. In contrast to the 
preceding example where the number of species was 
compared to changes in area size, the next situa- 
tion shows changes in species richness over time 
while the area remains constant (except for minor 
park boundary adjustments) : Mount Rainier National 
Park is located somewhat west of the central 
Washington Cascades. Mount Rainier is a large 
Pleistocene strato-volcano exceeding some 4000 
meters in elevation. The Park boundary encompasses 
about 97,000 hectares of forest, alpine tundra, 
glacier, and rock. The land on the east and south 
of the mountain is national forest land subject to 
multiple use management ranging from logging, gra- 
zing, mining, recreation, and all other comparable 

The northern boundary of the Park abuts a checker 
board of national forest and private lands in which 
alternative sections are owned by either the Weyer- 
haruser ("The Tree Growing") Company or the U.S. 
Forest Service. These alternately owned lands are 
managed jointly under a series of mutually bene- 
ficial agreements. The west Park boundary is 

8 1 


Land mammal distribution on the San Juan Islands and Vancouver Island. 

(321,000 ha) 

San Juan 
(53,400 ha) 

San Juan 
I si and 
(14,300 ha) 

(709 ha) 

Sorex obscurus 

S. palustris 

S. vagrans 
Plecotus townsendii 
Eptesucus fuscus 
Lasionycteri s noctivagans 
Corynorhinus townsendii 
Myotis lucifugus 

M. yumanensis 

M. evotis 

M. volans 

M. cal i fornicus 
Oryctolagus cuniculus 
Marmota vancouverensis 
Sciurus niger 
Tamiasciurus douglasi! 
Eutamias townsendii 
Castor canadensis 
Peromyscus maniculatus 
Microtus townsendii 
Ondatra zibethica 
Rattus norvegicus 

R. rattus 
Mus musculus 
Canls faml 1 iaris 
Vulpes fulva 
Ursus americanus 
Procyon lotor 
Martes americana 
Mustela erminea 

M. vison 

M. putorlus 
Lutra canadensis 
Gulo luscus 
Fel i s concolor 

F. domestlcus 
Cervus elaphas 
Odocoileus hemionus 

TOTAL (1975) 















Gowan and Guiquet 19 56. 
'Schoen 1972. 

TABLE 2. Turnover of mammalian species on San Juan Island, 1868-1975. 

1868 - 1925 


Immigrants (6) 
1868 - 1975 


Native Species (13) 

1868 - 1975 

Castor canadensis 
Canls lupus 
Ursus americanus 
Cervus elaphus 

Oryctolagus cuniculus* 
Rattus norvegicus 
Rattus rattus* 
Mus musculus 
Mustela putorlus* 
Fel Is domest icus* 

Sorex vagrans 
Epteslcus fuscus* 
Laslonycteris notivagans* 
Myotis yumanensis* 
Myotis luciflgus* 
Myotis cal i fornicus* 
Peromyscus maniculatus* 
Microtus townsendii 
Ondatra zibethica 
Procyon lotor* 
Mustela vison 
Lutra canadensis* 
Dama hemionus* 

* Present In San Juan Island National Historical Park 



J .2 3 4 SJ6JA910 2 3 4 5 6 78910 50 

FIGURE 1. A species-area curve for the San Juan Island, in which the number of species (ordinate) 
is plotted against the log of the area in thousands (10 ) of hectares (abscissa) . (Line slope is 
y = -16.78 + 8.25X.) SAJH is the position of San Juan Island National Historical Park. 

TABLE 3. Turnover of mammal species on Mount Rainier National Park. 

RATE: 1sp/3yrs 

I . Immigrat Ions 

A. 1920 - 1935 

P 1 ecotus townsend 1 1 
Erethizon dorsatum 
Gulo luscus 
Lutra canadensis 
Mephitis mephitis 

B. 1935 - 1976 
Laslurus clnereus 

I I . Turnover 

A. 1920 - 1935 

9 species In 15 years ■ lsp/1.67yrs 

B. 1935 - 1976 

8 species In k\ years - lsp/5.1yrs. 

1 1 . Ext Inct ions 

A. 1920 - 1935 

Myotis volans 
Rattus norveglcus 
Gulo luscus 
Lutra canadensis 

B. 1935 - 1976 

Scapanus townsend II 
Scapanus orarius 
Martes pennant i 
Spi logale putorlus 
Mephitis mephitis 
Canis lupus 
Lynx canadensis 

RATE: lsp/3-5yrs 

RATE: lsp/5.7yrs 











4 3% 


FIGURE 2. A figure showing the relative portion of land mammals west of the Cascade Mountains' 
crest in Washington State, by two authors (Taylor and Shaw 1920; Kitchen 1935) and this study. 

largely state and private timberlands. 

Early literature of the mammals of Mount Rainier 
National Park recorded 50 species present in the 
Park in 1920 (Taylor and Shaw 1927) . By the mid- 
1930 's, Kitchen (1935) reported 49 species could be 
found in the Park. The most recent study, which 
we have been conducting for the past 2 years veri- 
fies the presence of only 37 mammal species in the 
Park. This is an apparent extirpation of 13 (+ 2) 
species from Mount Rainier National Park since 1920 
(Figure 2) . We have checked as carefully as pos- 
sible to determine the reliability of the early 
reports. Fortunately, most of the specimens col- 
lected by Taylor and Shaw are present in the U.S. 
National Museum's Biological Survey collection, 

while Kitchen's records consist largely of anec- 
dotal observations by Park rangers and naturalists. 
By comparing these accounts with our own data, it 
is possible to estimate the insularity parameters 
of immigration and extinction rates (Table 3). 
Both immigration and extinction rates seem to be 
depressing over the past 40 years. This informa- 
tion suggests that perhaps relaxation is occurring 
and that a new equilibrium will be achieved at some 
future time; but at present, we simply do not have 
enough information to predict the final outcome, 
nor do we have enough to boldly state that the bio- 
geographic theories posed are truly applicable to 
the situation. 

8 6 

Recent controversies over the utility of bio- 
geographical notions to natural preserve design and 
management shed little light because the relevant 
data based for evaluating these ideas (and for eval- 
uating alternative positions) and their efficacy to 
national parks is lacking (See Simberloff and Abele 
1976; Diamond et al. 1976). If it can be demon- 
strated that the extinctions shown here are not a 
function of insularity, then alternative questions 
must be asked. Such alternatives must question 
those environmental influences which are known to 
disturb ecological systems and management practices 
or policies that may have led to the extinctions. 
Management programs in national parks often attempt 
to perpetuate a desired (static) situation in a 
natural system which is undergoing constant short- 
and long-term changes. These efforts are increas- 
ingly costly and in the long run, unlikely to suc- 
ceed. Pragmatic programs designed to facilitate 
equilibrium processes are more realistic, and they 
must be encouraged if the national parks are to pre- 
serve primeval habitat remnants for future genera- 
tions to experience. 


COWAN, I. McT., and C. J. GUIGUET. 1956. The mam- 
mals of British Columbia. British Columbia 
Provincial Museum Handbook II. Victoria. 414 p. 

DIAMOND, J. M. 1973. Distributional ecology of 
New Guinea birds. Science 179:759-769. 

DIAMOND, J. M. 1975. The island dilemma: lessons 
of modern biogeographic studies for the design 
of natural reserves. Biol. Cons. 7:129-146. 

LYNCH, P. A. OPLER, and C. S. ROBBINS. 1976. 
Island biogeography and conservation: strategy 
and limitations. Science 193:1027-1032. 

KITCHEN, E. A. 1935. Checklist of the vertebrates 
of Mount Rainier National Park. USDI National 
Park Service pamphlet. 

MacARTHUR, R. M. , and E. O. WILSON. 1967. The theory 
of island biogeography. Princeton Univ. Press, 
Princeton, N. J. 203 p. 

SCHOEN, J. W. 1972. Mammals of the San Juan Archi- 
pelago: distribution and colonization of native 
land mammals and insularity in three populations 
of Peromyscus manisulatus . M.S. Thesis. Univ. 
Puget Sound. 119 p. 

SIMBERLOFF, D. S., and L. G. ABELE. 1976. Island 
biogeography theory and conservation practice. 
Science 191:285-286. 

TAYLOR, W. P., and W. T. SHAW. 1927. Mammals and 
birds of Mount Rainier National Park. USDI- 
Natl. Park Serv. GPO, Wash. D.C. 249 p. 

WILSON, E. O., and E. O. WILLIS. 1975. Applied 
biogeography. in M. L. Cody and J. M. Diamond, 
eds. Ecology and evolution of communities. Belk- 
nap Press. Cambridge, MA. 



G. Abele 1 and Edward F. Connor 1 ' 


Dayton (1973) presented two examples of concep- 
tual ecological models whose prediction appeared 
to be verified by field observation, but upon clo- 
ser examination diametrically opposed mechanisms 
were actually indicated by the data. An analogy to 
Dayton's dilemma of "making the right prediction 
for the wrong reasons" has developed in attempts 
to apply island biogeography theory to the design 
of nature preserves. The right decision has been 
recommended, the preservation of maximum diversity, 
but the wrong strategy has been suggested. Further- 
more, Erhenfeld (1976) cautions that ecological 
theory is not always compatible with the desires 
of conservationists. 

The ethic of preserving maximum diversity is 
admirably and eloquently stated by Wilson and 
Willis (1976) and we wholeheartedly echo their sen- 
timents. However, we disagree with the specific 
design recommendations of Terborgh (1974, 1975), 
Diamond (1975), Diamond and May (1976), and Wilson 
and Willis (1976) , and show here that they have pre- 
sented the wrong reasons for the right decision. 
These authors essentially contend that specific de- 
sign recommendations regarding the size and shape 
of nature preserves can be derived from island bio- 
geography theory, and that these recommendations 
represent the design strategy capable of preserv- 
ing maximum diversity. They proceed to build a 
case for the construction of single large nature 
preserves as opposed to multiple small preserves, 
and make further recommendations concerning the 
shape of nature preserves. Two classes of evidence 
are presented in support of their recommendations; 
(1) arguments based on the species-area relation- 
ship, and (2) arguments based on the equilibrium 
theory of island biogeography, especially the phe- 
nomenon termed "relaxation." We offer here a pre- 
liminary examination of this complex problem and 
suggest that application of theory to conservation 
strategy is not as well supported by data as the 
literature would suggest. 


The species-area relationship is an empirical re- 
lationship that is consistent with the equilibrium 
theory of island biogeography (Simberloff 1974) , 
but was derived long before and independent of 
island biogeography theory. The species-area rela- 
tionship is simply a formalization of the observa- 
tion that large areas usually contain more species 
than small areas. Three mechanistic explanations 
of the species-area relationship have been proposed, 
two prior to the advent of modern island biogeo- 
graphy theory, and one a prediction of equilibrium 

Habitat diversity has been the most popular ex- 
planation of the species-area relationship. It has 
long been proposed that species numbi rs increase 
with area because larger areas have more habitats 

Department of Biological Science, Florida State 
University, Tallahassee 32306. 

Tall Timbers Research Station, Route 1, Box 
160, Tallahassee 32303. 

and therefore species peculiar to particular habi- 
tats are more likely to be encountered when exam- 
ining large areas. An alternative to the "habitat 
diversity hypothesis," also proposed prior to 
equilibrium theory yet based on island observations, 
is that islands passively sample a dispersing biota, 
larger islands intercepting a larger sample of in- 
vaders and hence a larger number of species. As 
a subsidiary prediction of equilibrium theory a 
third hypothesis was proposed; species numbers 
increase with area because species on larger 
areas (or islands) have on the average larger 
population sizes, and therefore a lower chance of 
becoming locally extirpated, so that at any one 
time more species will coexist locally on large 
islands than on small islands. 

As often occurs in ecology, several factors 
may play a causal role, and good evidence exists 
that each of these three mechanisms contributes 
to the species-area relationship (Abele 1974; 
Harmon 1972; Simberloff 1976a; Abele and Patton 
1976; Osman, 1977) . However, biogeographers 
have as yet been unable to determine the 'pro- 
portional contribution' of these three mechanisms 
to the observation of a positive correlation be- 
tween species number and area. 

Graphically the species-area relationship has 
been represented in two forms; (a) with species 
numbers and areas untrans formed (Figure 1) , and 
(b) with species numbers and areas both logarith- 
mically transformed (Figure 2) . The curve in 
Figure 1 represents the usual shape of the re- 
lationship between species number and area, and 
is described by the equation 

S = kA z , 
where S equals species number, A equals area, and 
k and z are constants. However, when species 
numbers and areas are examined in their untrans- 
formed modes data are fitted to the equation 
S = zA+k. The solid line in Figure 2 represents 
the same relationship with species numbers and 
areas logarithmically transformed. The line is 
both described by and commonly fitted to the equa- 
tion logS = zlogA+logk. In both instances, the 
positive slope of the curves indicates the empiri- 
cal result that species numbers are larger for 
large areas. 

The major recommendation made for conservation 
practice and the design of biotic preserves, pur- 
portedly based on island biogeography theory, is 
that the largest single land area obtainable will 
preserve the greatest number of species and is 
therefore the best strategy (Terborgh 1974, 1975; 
Diamond 1975; Diamond and May 1976; Wilson and 
Willis 1975) . However, this recommendation is 
predominantly based on the species-area relation- 
ship and is nothing more than an attempt to trans- 
late the long known monotonically increasing re- 
lationship between species number and area into a 
design strategy. Actually the species-area re- 
lationship is ambiguous on this point. Given that 
preserving the largest number of species possible 
is the goal to support for two conflicting strat- 
egies; (a) a sinqle large preserve or (b) multi- 
ple smaller preserves with a combined area equal 
to the single large preserve. 

The fact that species number is a monotonically 
increasing function of area supports the observa- 


tion that a large preserve will contain more species 
than a smaller one. However, the exact shape of 
the functional relationship between species num- 
ber and area argues that Strategy b (multiple small 
preserves) may actually preserve more species. In 
Figure 1 the curve depicting the relationship be- 
tween species number and area becomes asympotic 
(levels off) with large areas. Fewer and fewer 
species are added per unit area as larger areas 
are examined. In other words, the per unit area 
return in species preserved diminishes rapidly for 
large areas. The same pattern is revealed in Fig- 
fure 2 when log-species numbers and log-areas are 
plotted. The sloping dot-dash line represents a 
species-area relationship with a slope value of 
z = 1. At this slope value the return in species 
preserved per unit area is constant regardless of 
the absolute size of the area. Species-area re- 
lationships with slope values greater than 1 indi- 
cate that the per unit return in species preserved 
increases with large areas, and species-area re- 
lationships with slope values less than 1 indicate 
a decreasing return in species preserved per unit 
area with increasing area. Connor and McCoy 
(1979) have calculated from 100 species-area curves 
including multifarious taxa (e.g., birds, insects, 
plants, mammals, etc.) that the average slope value 
from actual observations is z = 0.3; much less than 
the value of z = 1 needed to maintain at least an 
equal return in species preserved per unit area. 
It would appear that in almost all natural situa- 
tions the per unit area return in species pre- 
served will always be much less for a large refuge 
than for a smaller one. 

Related to this probl 
is a further ambiguity o 
tionship regarding conse 
ample, if a preserve is 
likely to have more spe 
area B = >sA. However, a 
(1962) Simberloff (1974 
(1976; but see comments 
1976; Whitcomb et al. 19 
lationship predicts that 
serves each of area B = 
serves together will con 

em of diminishing return 
f the species-area rela- 
rvation practice. For ex- 
built with area A it is 
cies than a preserve of 
s pointed out by Preston 
and Simberloff and Abele 
by Diamond 1976; Terborgh 
76) the species-areas re- 

if one builds two pre- 
^A, the two smaller pre- 
tain more species than the 


FIGURE 1. Plot of the species-area relation- 
ship for untransformed species numbers and areas, 
illustrating the asymptotic behavior of species 
number for large areas. 

single large preserve. Figure 2 graphically illu- 
strates this situation. Consider preserve A with 
S A species and area A, and a smaller preserve B 
with Sb species and area B = *5A. Note that the 
distance between the points S ft and S B is smaller 
than the distance between the points A and B. This 
indicates that preserve B with half the area of pre- 
serve A will contain more than half the species of 
preserve A, so that two preserves of area B will 
contain more species than preserve A, but comprise 
an equal land area. In essence, the species-area 
relationship predicts that two small preserves will 
contain more species than one large preserve. Fig- 
ure 3 shows exactly what proportion of the species 
of a particular area will be preserved given that 
a certain proportion of the area is maintained as 
a refuge. In the example above, 50% of the land 
area preserves 81% of the species, and if only 10% 
of the land area were to remain still 50% of the 
species would be preserved. 

We emphasize that although two small preserves 
are predicted, based on the species-area relation- 
ship to contain more species than a single large 


FIGURE 2. Plot of the species-area relationship for logarithmically transformed species- 
numbers and areas. The dot-dash line represents a species-area relationship with a slope 
value of Z = 1, indicating a constant return in species number independent of the absolute 
area. The solid line with a slope of Z = 0.3 indicates the usual situation found in nature. 
Point A represents a nature preserve with area A, and point B a preserve with area B = ^A. 
The dasned lines project these areas onto the species-area curve with Z = 0.3 and then to 
the vertical axis so that the number of species contained by each preserve can be determined 
(points S^ and S B ) . The small distance between points A and B, indicates that preserving 
half the area will preserve more than half the species, so that two small preserves of area 
B = HA may contain more species than preserve A. 


one of equal total area, we cannot predict a priori 
that the two small preserves will contain more spe- 
cies than a single large one since the two small 
preserves could potentially share many species. 
Realistically we may expect the two small preserves 
to share at least some species and to each possess 
some unique species, but without knowledge of the 
habitat requirements of the species considered it 
is impossible to determine which is the superior 
strategy, a single large or multiple small pre- 

A final comment on the implications of the spe- 
cies-area relationship for conservation practice 
is that the species-area relationship does not ex- 
plain species numbers equally well in all regions. 
Based on 100 species-area curves (log-log), Connor 
and McCoy (1979) determined that the parametric 
correlation coefficient of the species-area rela- 
tionship was significantly negatively correlated 
with latitude (r = -0.3181, p = 0.0001). 
This indicates that the per unit return of spe- 
cies preserved decreases from the tropics to the 
temperate zone. This further points out that the 
correct design formula for biotic preserves is 
different for each specific case. 


The equilibrium theory of island biogeography 
has been applied to the design and maintenance of 
parks and refuges. Its basic premise is that the 
biota of an island is in dynamic equilibrium be- 
tween the immigration of species new to the is- 
land and local extirpation of those already present. 
Thus the total number of species remains about the 
same, but the composition changes over time. This 
change in composition without a change in species 
numbers is known as 'equilibrium turnover.' In 
assessing the occurrence and importance of equi- 
librium turnover we must be concerned with the ex- 
tinctions of previously persistent, reproducing 
populations that disappear independent of secular 
environmental changes, specifically human influence 
(Lynch and Johnson 1974; Simberloff 1976a). A corol- 
lary to equilibrium turnover is the phenomenon of 
'non-equilibrium turnover. ' Non-equilibrium turn- 
over is the result of an imbalance in extinction and 
immigrations and is attributable to secular environ- 
mental changes. In the case of parks and refuges it 
is usually caused by human influence. During non- 
equilibium turnover, changes in species numbers 
and compositional changes occur simultaneously. 

20 40 60 80 tOO 

Decrease in Area (%) 

FIGURE 3. A plot of the percentage decrease in 
species number which accompanies a percentage de- 
crease in area. For example, a decrease of 90% 
in area results in a loss of 50% of the species. 
The figure is based on S = kA z where Z = 0.3. 

The implication of equilibrium turnover for the 
design of nature preserves is apparent in that if 
equilibrium turnover is a significant phenomena, it 
would indeed be very difficult to design a refuge 
for any particular species or group of species 
since the composition of the refuge's biota would 
be expected to change constantly. If such were 
the case, designing preserves to increase species' 
persistence would be paramount. The implication 
of non-equilibrium turnover for preserve design is 
that the use and management both of parklands and 
surrounding regions can be expected to affect the 
number and composition of species that a preserve 
will contain. This is hardly a novel message, and 
anything but uniquely predicted by equilibrium 
theory. However, we are concerned here with the 
assessment of the significance of equilibrium 
turnover and the direction, magnitude, and time 
scale of non-equilibrium turnover (especially re- 
laxation, species loss) caused solely by change in 
island's or a preserve's area and insularity. 

Equilibrium Turnover 

There appear to be four sets of data that dem- 
onstrate turnover at equilibrium (1) arthropods on 
mangrove islands in the Florida Keys (Simberloff 
and Wilson 1969; Simberloff 1976a, 1976b) (2) birds 
of the Channel Islands off California (Diamond 
1969) (3) birds of Karkar Island off New Guinea 
(Diamond 1971) (4) birds of islands off Britain 
and Mona Island off Peurto Rico (Terborgh and 
Faaborg 197 3) . 

Arthropods on Mangrove Islands 

In the experimental test of equilibrium theory 
Simberloff and Wilson (1969) measured turnover 
rates of approximately 0.5 species/island/day or 
182 species/island/year. Such a high extinction 
rate would be cause for serious concern among 
those interested in refuge planning. However 
Simberloff (1976b) has carefully reexamined the 
data and suggests that much of the turnover ob- 
served was due to "pseudoturnover" or normal 
inter-island movement by transients and did not 
involve local extinction of long-term reproducing 
populations. Using conservative criteria Simber- 
loff (1976a) estimates a minimum extinction rate 
of 1.5 species/island/year, a considerable re- 
duction from the previous estimate of 182 species/ 
island/year. Although an extinction rate of 1.5 
species/year would be significant for a nature pre- 
serve, these values, computed from the mangrove 
island arthropod fauna, are not indicative of 
what can be expected of other taxa (birds, mam- 
mals, etc.) or the same taxa in other environments. 
Mangrove islands, situated on the constantly 
changing shoreline, certainly provide a more ephem- 
eral environment (hence higher extinction rates) 
than the analogous upland forest does for its 
arthropod inhabitants. 

Birds on Islands 

Estimates of turnover rates for the avifauna of 
islands have been obtained by comparing species 
lists from surveys taken at different times. The 
assumptions have been that the islands are at 
equilibrium and that a changed composition with 
no change in species number is indicative of 
equilibrium turnover. For example, Diamond (1969) 
compared surveys of the birds of the Channel Islands 
taken in 1917 and 1968 and concluded that the is- 
lands were in equilibrium and that between 17 and 
62% of the breeding species has been lost (and 
replaced by other species) over the 51-year period. 
Similar figures are reported by Diamond (1971) and 
Terborgh and Faaborg (1973) . Diamond and May (19 76) 
summarize these studies reporting turnover rates 
(per year) of 0.2 to 20% of the island's bird spe- 
cies for islands of 0.4 to 400 km^. This corres- 
ponds to approximately 0.1 to 0.3 species/island/ 


year. The critical questions relevant to these 
studies are: Did the extinctions involve species 

which had previously persisted and reproduced on 
the islands and were the extinctions independent 
of human influence? Lynch and Johnson (1974) have 
addressed these questions with special reference 
to Diamond's (1969) report on the Channel Islands. 
They were able to avaluate 41 of the 48 claimed 
extinctions and concluded that 33 (80%) were 
either related to human activities or were the 
result of "pseudoturnover. " They believe that of 
the realining species only two represent valid ex- 
tinctions of reproducing populations. This would 
lower the estimated extinction rate from 0.10 
species/island/year to about 0.004 species/island/ 
year. While it is conceivable that Lynch and 
Johnson's conservative approach may have excluded 
some valid extinctions we are in general agree- 
ment with their conclusion that extinction rates 
have been greatly overestimated. 

Rel axa ti on 

Situations exist, particularly with respect to 
parks and refuges, where the initial area of forest 
or habitat decreases. Since species number is re- 
lated to area, some change in species number can 
be expected (Figure 3) as already discussed. This 
process of reequilibration has been referred to 
as "relaxation" by Diamond (1972) . The rate of 
species loss (extinction rate) generated by this 
process of insularization is important to park 
and refuge design in two ways. First, if extinc- 
tion rates are related to the area of the preserve 
then the nature of the relationship may be germane 
to the question of single large or multiple small 
preserves. For example, if extinction rates are 
higher for small islands the persistence of spe- 
cies on small islands will be less than on large 
islands, and large islands would make better nature 
preserves. Second, what is the time scale over 
which extinctions caused by changing the area or 
more completely insularizing a preserve cover? 
If the time scale is short, we know that the activ- 
ities in regions surrounding parks (e.g., urbani- 
zation and clear-cutting) will have an immediate 
effect on the diversity of species in parks. If 
the time scale is long, then it may be possible 
to allow short term high-yeild exploitation of 
these resources and make amends (e.g., reforesta- 
tion) later, before extinctions occur. 

Three data sets are pertinent to these questions; 
Diamond's (1972, 1975) analysis of relaxation times 
for the avifauna of islands off New Guinea, Ter- 
borgh's (1974, 1975) analysis of the West Indian 
and Barro Colorado Island avifaunas, and Simber- 
loff's (1976b) analysis of the arthropod fauna of 
mangrove islands. 

Arthropods on Mangrove Islands 

The relationship between island size and aspects 
of island biogeography has been discussed by Simber- 
loff (1976a) and here we only summarize and com- 
ment on his results. A decrease in the size of 
mangrove islands resulted in a decrease in spe- 
cies numbers. However, there was no clear rela- 
tionship between extinction rate and island size. 
Three rates rose and two declined with a decrease 
in area. Reequilibration from an oversaturated 
condition was rapid, all nine islands had reequi- 
librated within a year. 

Land Bridge Is lands and Birds 

Diamond (1J72, 1975), Diamond and May (1976) 
and Terborgh (1974, 1975) estimate relaxation times 
and extinction coefficients for islands avifaunas 
using land bridge islands. The assumption is that 
islands separated from the source fauna by water 

less than 100 meters in depth were connected to 
the mainland (or source) about 10,000 years ago 
when sea levels were much lower. Thus these is- 
lands have experienced a change in area and pre- 
sumably a change in species number. The original 
avifauna presumed to have been present is esti- 
mated and compared to both the species number 
present now and the species number predicted by the 
species-area relationship for that region. We 
have examined all relevant data (Abele and Connor, 
in prep.) and can only summarize our findings 
here. Diamond (1972, 1975) estimates the original 
species number for the land bridge islands to be 
325, the entire lowland avifauna of New Guinea. 
This is biologically and statistically unrealist- 
ic, for as Diamond (1972, 1975) himself emphasizes, 
the distribution of New Guinea birds is extremely 
patchy and the probability of 325 species' breed- 
ing in an area the size of Misol (2040 km^ com- 
pared to New Guinea 319,713 km^) must approach 0. 
Similar comments apply to Terborgh 's estimates. 
Both Diamond and Terborgh note that land bridge is- 
lands tend to fall above a species-area regres- 
sion line for the respective regions. They cite 
this as evidence that the islands are 'supersat- 
urated' with species. We have reexamined the 
New Guinea data and calculated a species-area 
regression for 67 islands (both with and without 
New Guinea included) . If the land bridge islands 
are in fact supersaturated, one would expect the 
regression residuals to be significantly greater 
than the mean residual. Not a single island, in- 
cluding all land bridge islands has a residual 
three or more standard deviations from the mean, 
the usual criterion for outliners (Draper and 
Smith 1966) . Only four islands are two standard 
deviations from the mean, one is negative and 
three are positive. These are Ritter (negative) , 
already noted by Diamond to be vegetatively de- 
pauperate, a coral islet (positive) previously sug- 
gested by Diamond to be species poor (Diamond 1974) , 
and two land bridge islands, Salawati and Botanta 
(both positive) . It is true that land bridge is- 
lands tend to be above the regression line but this 
may be expected since they are close to the source 
fauna. Thus there is no evidence that New Guinea 
land bridge islands are supersaturated. We will 
deal with the West Indian islands elsewhere but 
Terborgh's (1975) calculation of extinction co- 
efficients and their supposed inverse correla- 
tion with area deserve comment. It can be shown 
(Abele and Connor, in prep.) by simple algebraic 
substitution that Terborgh's formula for K-, (ex- 
tinction coefficient) reduces to K3ac 1/A (since 
the other terms in the right hand expression are 
all constants) . Therefore his regression of K3 
against area is circular and the conclusion that 
extinction rates decrease with increasing area is 
trivial and tautological. 

Barro Colorado Island 

Barro Colorado Island deserves special comment 
because it is the only island for which exist ac- 
tual census of the original birds present. Willis 
(1974) and Wilson and Willis (1976) reviewed the 
data and concluded that of 209 original species 
45 had become extinct over a 50-year period. They 
note that most (32) of these can be explained to 
be the result of habitat changes but the reasons 
for the other 13 remain elusive. Terborgh (1974, 
1975) estimates an extinction coefficient for the 
island and a predicted loss of 16-17 species. This 
result has received wide acclaim as " . . .an 
acceptable account of reality . . . . " (Terborgh 
1975) ". . .gratifyingly in accord with the 
actual . . .." Diamond and May 1976) and "... dra- 
matically confirmed the accuracy of his calcula- 
tions . . . ." (Diamond 1975). Disregarding for 
the moment the mathematical circularity involved 
in the calculations (Abele and Connor, in prep.) 
we find (based on Terborgh's own data) that 95% 


confidence interva 
We cannot agree th 
of reality. Our c 
shaken by an exami 
numbers present on 
(1938) and Eisenma 
of which 253 they 
dence interval aro 
209 species must a 

1 for the estimate to be 7 to 81. 
at this is an acceptable account 
onfidence in the data is further 
nation of the original species 

Barro Colorado. Since Chapman 
nn (1952) list about 300 species, 
consider resident, the confi- 
und Willis's (1974) estimate of 
lso be fairly wide. 

Time Scale for Relaxation 

The relaxation times suggested by Diamond and 
Terborgh are on the order of 10,000 to 20,000 
years. Is this reasonable? We have already noted 
that the arthropods of mangrove islands reached a 
new equilibrium level in less than a year. Regard- 
less of the ultimate causes it is clear that major 
vegetational and avifaunal changes occurred on 
Barro Colorado in less than 50 years. Stronq (1974a, 
1974b) and Strong, McCoy and Rey (1977) have 
presented persuasive arguments that phytophagous 
insects equilibrate on introduced host plants with- 
in a few hundred years. This involves host plants 
switching, a process thought to require evolution- 
ary time. We would suggest that whatever species 
loss may have occurred on land bridge islands, it 
occurred rapidly, lagging behind the actual sea 
level change very little. 

The low values of equilibrium turnover calcula- 
ted by Lynch and Johnson (1974) and Simberloff 
(1969) suggest that it has little significance for 
the design of nature preserves. Equilibrium theory 
predicts that extinction rates should be inversely ' 
correlated with area, but the experimental work of 
Simberloff (1976a) yielded no pattern. The inverse 
relationship between extinction rates and area re- 
ported by Terborgh (1974) for several West Indian 
Islands is based on circular mathematical reason- 
ing and is therefore uninformative. The time scale 
of relaxation reported by Terborgh (1974, 1975) and 
Diamond (1972, 1975) is extremely long. However, 
the analyses of these authors are beset with many 
statistical problems, and are based only on infer- 
ences about original species numbers. Simberloff s 
(1976a) experiments suggest that the time scale of 
relaxation is very short. 


A major problem in designing nature preserves 
is to determine a priori which species will be main- 
tained. Island biogeography theory assumes that 
species with low abundances are more likely to be 
lost than common species. However, the presence or 
absence of a species from a particular preserve 
cannot be predicted simply by a knowledge of its 
abundance before the preserve is developed, but 
must include an in depth knowledge of the biology 
of the species in question. For example, low abun- 
dance "fugitive" species are extremely persistent 
when presented with a patchy environment. 

Some empirical observations generally classed 
under the rubric of 'ecological truncation' are 
pertinent to this problem. Ecological truncation 
refers to the loss of specialists and of large 
members of ecological guilds. For example, there 
were originally 11 or so species of ant- following 
birds on Barro Colorado Island (Willis 1974) of 
which three became extinct or nearly so during 
1960-70. These three were the largest members of 
the ant- following guild. Pelicans, ospreys, and 
eagles are three other examples. Large-bodied ani- 
mals usually have large feeding territories (McNab 
1963; Schoener 1968), as to the so-called top 
carnivores. These animals are extremely vagile 
and are likely to move in and out of even the 
largest nature preserves. The disappearance of the 
elk, wolves, and bear from Mount Rainier National 
Park is a perfect example of this problem (Weis- 


brod, this volume) . 

Somewhat similar to the absence of large- 
sized animals is the often reported absence of 
certain families of birds from islands. Mac- 
Arthur, Diamond and Karr (1972) for example note 
the absence of certain families of birds from the 
Pearl Islands on the Pacific coast of Panama and 
conclude that it is unlkely to be accidental. 
However, only rarely (Simberloff 1974) has the 
question been asked, "What is the probability of 
these families being represented in a random 
sample of say, 46 species (e.g., Rey Island) of 
Panamanian birds?" For example, of 20 random draws 
of 59 species from the Panamanian avifauna 17 
lacked families containing 20 or more species and 
three times families with 30 or more species were 
missed (Simberloff 1974) . Only after the problems 
of random sampling probabilities have been factor- 
ed out can the interesting ecological patterns be 

The emphasis on ecological truncation, the 
absence of certain families, and local extirpation 
have all been framed around islands without refer- 
ence to these phenomena in mainland regions. For 
example the Harpy Eagle is no longer present on 
Barro Colorado (ecological truncation and local 
extirpation) but what is the probability of not 
only finding the species but finding it breeding 
in any 15.6 km^ region of the adjacent mainland? 
How often do mainland species disappear from any 
subset of the region? What is the probability of 
finding jacamars, puff birds and guans in any small 
region of the mainland? Studies such as those of 
Weisbrod (this volume) , Moore and Hooper (1975) , 
and Galli et al. (1976) can help to answer these 
questions. However, a more complete and sophisti- 
cated knowledge of species' abundances, regional 
movements, and habitat requirements are the only 
data capable of yielding the necessary predictive 
power to permit the effective and efficient de- 
sign of nature preserves. 

The previous discu 
a very risky business 
priori the single bes 
to preserve maximum d 
studies of the habita 
floras and faunas. I 
conclude that a singl 
effective in preservi 
multiple small preser 
alities of the land a 
potential for placing 
habitats are consider 


ssions have shown that it is 

to attempt to predict a 
t design strategy in order 
iversity without detailed 
t requirements of entire 
t is at best premature to 
e large preserve will be more 
ng maximum diversity than 
ves, especially when the re- 
cquisition process and the 

smaller preserves in varied 

The design criteria of Terborgh (1974, 1975), 
Willis (1974) , Diamond (1975), Wilson and Willis 
(1975) and Diamond and May (1976) all stress a 
single large area as being preferred for theoret- 
ical reasons. Implicit in the arguments in favor 
of a single large preserve is the connotation that 
small preserves are of less use. This is because 
most writers have dealt with vertebrates, particu- 
larly birds and mammals. A 1 km^ refuge might be 
rejected as inappropriate for birds and mammals 
but it might be fine for some rather magnificant 
spiders or soil arthropods. Biologists know little 
of the population dynamics of vertebrates and al- 
most nothing of invertebrates. We would not reject 
any amount of land from consideration as a refuge. 
As pointed out, arguments based on the species- 
area relationship are ambiguous on this point, and 
the extremely long relaxation times calculated by 
Terborgh and Diamond actually argue that we need 
not act with haste to forestall extinctions. The 
careful experiments of Simberloff (1976a), how- 
ever, show that relaxation times for one fauna are 
rapid and that action may be needed now to prevent 
extinctions . 

We also point out that reaso 
gestions must distinguish betw 
roles for nature preserves; th 
local extirpations, and the pr 
extinctions. Approaches to th 
recognize both the various des 
that nature preserves may be e 
There is value in preventing 1 
availing the local biota to re 
but on the periphery of a spec 
dubious that disproportionate 
made to preserve local populat 
efforts well within the distri 
a species would be much more e 

nable design sug- 
een two possible 
e prevention of 
evention of ultimate 
is problem need to 
ign goals and uses 
xpected to fulfill, 
ocal extirpation, 
sidents of a region, 
ies ' range it is 
efforts should be 
ions when lesser 
butional range of 
f f ective. 


We wish 
of organic 
much as po 
as little 
is with th 
of an insu 
large body 
Service ca 
the theory 
many of th 
five to te 
benefit ev 
Service is 
able natio 

to reaffirm our support of the ethic 

diversity and urge that mankind do as 
ssible to preserve organic diversity and 
as possible to disturb it. Our complaint 
e uncritical acceptance and application 
fficiently validated hypothesis. The 

of empirical observations which support 
on practices remains a firm foundation 
t of any theory. The National Park 
n contribute to both the data base and 

by continuing to support and enlarging 
e monitoring programs discussed at this 

Many of the questions relating bio- 
and conservation can be probed by simply 
populations of animals in parks over a 
n year period. Such projects would 
eryone concerned. The National Park 

guardian of a precious and irreplace- 
nal resource and this bears a special 
lity (and burden) to safeguard it. 


ABELE , L. G. 1974. Species diversity of decapod 
crustaceans in marine habitats, Ecology 55: 

ABELE, L. G., and W. K. PATTON . 1976. The size of 
coral heads and the community biology of asso- 
ciated decapod crustaceans. J. Biogeo. 3:34-47. 

CHAPMAN, F. M. 1938. Life in an air castle. D. 
Appelton-Century Company 250 p. 

CONNOR, E. F. and E. D. McCOY . 1979. The statis- 
tics and biology of the species-area relation- 
ship. Amer. Nat. in press. 

DAYTON, PAUL K. 1973. Two cases of resource parti- 
tioning in an intertidal community: Making the 
right prediction for the wrong reason. Amer. 
Nat. 107:662-670. 

DIAMOND, J. 1969. Avifaunal equilibria and species 
turnover rates on the Channel Islands of Cali- 
fornia. PNAS 64:57-63. 

DIAMOND, J. 1971. Comparison of faunal equilibrium 
turnover rates on a tropical island and a tem- 
perate island. PNAS 68:2742-2745. 

DIAMOND, J. 1972. Biogeographic kinetics: Estima- 
tion of relaxation times for avifaunas of south- 
west Pacific islands. PNAS 69: 3199-3203. 

DIAMOND, J. 1974. Colonization of exploded volcan- 
ic islands by birds: The supertramp strategy. 
Science 184:803-806. 

DIAMOND, J. 1975. The island dilemma: Lessons of 
modern biogeographic studies for the design of 
natural reserves. Biol. Conservation 7:129-146. 

DIAMOND, J. 1976. Island biogeography and conser- 
vation: Strategy and limitations . Science 193: 

DIAMOND, J. and R. MAY. 1976. Island biogeography 
and the design of natural reserves. in R. M. 
May (ed) Theoretical ecology principles and 
applications. W. B. Saunders Co., Philadelphia, 
PA. 163-186. 

DRAPER, N. R., and H. SMITH. 1966. Applied regres- 
sion analysis. John Wiley & Sons, Inc. 407 p. 

EHRENFELD, D. W. 1976. The conservation of non- 
resources. Amer. Sci. 64:648-656. 

EISENMANN, E. Annotated list of birds of Barro 
Colorado Island, Panama Canal Zone. Smithson- 
ian Miscellaneous Collections 11 7 (5) -. 1-62 . 

GALL I , ANNE E., C. F. LECK, and R. T. T. FORMANN . 
1976. Avian distribution patterns in forest 
islands of different sizes in central New 
Jersey. Auk 93:356-364. 

HARMON, W. N. 1972. Benthic substrates: their ef- 
fect of fresh-water mollusca. Ecology 53:271-277. 

LYNCH, J. F., and N. K. JOHNSON. 1974. Turnover 

and equilibria in insular avifaunas with special 
reference to the California Channel Islands. 
The Condor 76:370-384. 

McARTHUR, R. , J. DIAMOND, and J. KARR . 1972. Den- 
sity compensation in island fauna. Ecology 53: 

McNAB, B. K. 1963. Bioenergetics and determination 
of home range size. Amer. Nat. 97:133-140. 

MOORE, N. W. , and M. D. HOOPER. 1975. On the number 
of bird species in British woods. Biol. Conserv. 

OSMAN, R.W. 1977. The establishment and develop- 
ment of a marine epifaunal community. Ecological 
Monographs. 47:37-63. 

PRESTON, F. W. 1962. The canonical distribution of 
commonness and rarity I. II. Ecology 43:185-215, 

SCHOENER, T. 1968. Sizes of feeding territories 
among birds. Ecology 49:123-141. 

SIMBERLOFF, D. 1974a. Equilibrium theory of island 
biogeography and ecology. Ann. Rev. Eco. Syst. 

SIMBERLOFF, D. 1974b. Models in biogeography. in 
T. J. M. Schopf (ed) . p 160-191. Models in 
Paleobiology. Freeman, Cooper & Company, San 
Francisco, California. 

SIMBERLOFF, D. 1976a. Experimental zoogeography 
of islands: Effects of island size. Ecology 

SIMBERLOFF, D. 1976b. Species turnover and equili- 
brium biogeography. Science 194:572-278. 

SIMBERLOFF, D., and L. G. ABELE. 1976. Island bio- 
geography theory and conservation practice. 
Science 191:285-286. 

SIMBERLOFF, D., and E. WILSON. 1969. Experimental 
zoogeography of islands: The colonization of 
empty islands. Ecology 50:278-296. 

STRONG: D. R. 1974a. Non-asymptotic species rich- 
ness models and the insects of British trees. 
PNAS 71: 2766-2769. 

STRONG, D. R. 1974b. Rapid asymptotic species ac- 
cumulation in phytophagous insect communities: 
The pests of Cacao. Science 185:1064-1066. 

STRONG, D. R., E. McCOY, and J. REY . 1977. Time 
and the number of herbivore species: the pests 
of sugarcane. Ecology 58:167-175. 

TERBORGH, J. 1974. Preservation of natural diver- 
sity: The problem of extinction prone species. 
Bioscience 24:715-722. 

TERBORGH, J. 1975. Faunal equilibrium and the de- 
sign of wildlife preserves p. 369-380. in F. B. 
Golly and E. Medina (eds) Tropical ecological 
systems. Trends in terrestrial and aquatic re- 
search. Springer-Verlag: New York Ecological 
Studies Vol. 11. 

TERBORGH, J. 1976. Island Biogeography and conser- 
vation: Strategy and limitations. Science 193: 

TERBORGH, J., and J. FAARBORG . 1973. Turnover and 
ecological release in the avifauna of Mona 
Island, Puerto Rico. Auk 90:75-79. 

1976. Island biogeography and conservation: 
Strategy and limitations. Science 193:1030-1032. 

WILLIS: E. O. 1974. Populations and local extinc- 
tions of birds on Barro Colorado Island, Panama. 
Ecological Monographs 44:153-169. 

WILSON: E., and E. O. WILLIS. 1975. Applied bio- 
geography. in M. L. Cody, p. 522-534. and 
J. M. Diamond (eds) Ecology and evolution of 
communities. Harvard Univ. Press. Cambridge, 
Mass . 





Larry E. Morse and Jane I. Lawyer 

To be an effective interpreter of natural 
wonders and protector of diverse public lands, 
the U. S. National Park Service needs reliable, 
scientifically reviewed floristic information of 
many kinds. Traditional requirements for species- 
related knowledge about plants include specimen 
identification, contributions to interpretive pro- 
grams, and interdisciplinary needs of anthropolo- 
gists, historians, and paleontologists. The non- 
traditional questions raised in enlightened 
management of endangered populations and irreplace- 
able ecosystems also require answers based on 
reliable botanical data. At present, these diverse 
needs are partially fulfilled by regional floras 
and manuals, taxonomic monographs, ecological 
studies, and other traditional scientific works. 
However, the quality of such reference works 
varies considerably from region to region and 
taxon to taxon, and the information contained in 
them has never been synthesized satisfactorily 
and reviewed on a national scale. 

Several years ago, the Flora North America 
project (based at the Smithsonian Institution) 
began such a synthesis, but this work was termi- 
nated abruptly by austerity budgets in 1973. 
Flora North America proposed as twin goals the 
production of a published floristic treatise on 
the vascular plants of North America north of 
Mexico, and, concurrently, the development of a 
computer-based information resource including 
these data and associated information (Shetler, 
1971) . A preliminary checklist was distributed, 
and contributions were solicited from hundreds of 
plant systematists, but work was suspended before 
any of these contributions could be completed, 
edited, and included in the data base. 

Efforts are now being made to resume a modified 
Flora North America project with Department of 
the Interior sponsorship, as part of the Man and 
the Biosphere Program. The New York Botanical 
Garden is conducting a contract study in this 
area for the National Park Service, requiring in- 
vestigation of the floristic information needs of 
the National Park Service and other selected land- 
management agencies, to determine which of their 
information needs could be served through resump- 
tion of a computer-based flora project. The 
following discussion presents some of the prelimi- 
nary findings of the study. 

First, a computer-based floristic information 
system would provide considerable support to sci- 
entific research in National Park areas, whether 
conducted by Park Service staff or by independent 
scientists. There are many lacks and ambiguities 
in the species-related botanical information 
available for the United States; if this knowledge 
were organized into a systematic review of con- 
tinental scope, appropriate topics for future 
studies would be more evident. Meanwhile, great 
care would be needed to assure that the informa- 
tion-system design recognized the uneven quality 

and varying reliability of the available data 
(cf. Watson, 1971; Shetler, 1974; Morse, in 
press) . This systematized floristic information 
base, if cautiously used, could also support mod- 
ern information-retrieval procedures such as 
computer-based specimen identification techniques 
(Morse, 1975) . Users of the computer system 
would have to realize, however, that such methods 
are only as good as their taxonomic reference data, 
and that it will take many years of research to 
achieve consistently high quality. Nevertheless, 
a modern continental synthesis of floristic infor- 
mation is an overdue task which would enhance the 
productivity of new research in many areas (Gates, 
1971; Irwin, 1973; Raven, 1974) . 

The topic of legal-protection status of various 
taxa in various jurisdictions was not considered 
in the original FNA plans, but now deserves 
attention as an area where a centralized, computer- 
based information service could offer latest- 
available status reports to a large user community. 
Such a data base could be developed independently 
of Flora North America , but its coordination with 
FNA would provide a high quality scientific review 
particularly in the highly technical area of tax- 
onomic synonymy. Another example of an area in 
which some coordination with FNA seems desirable 
is taxon-related management information (such as 
fire ecology, reproductive strategies, or sensi- 
tivity to feral animals); again, wide dispersion, 
up-to-date revision, and quality review of the 
information are important considerations. 

The selection of new park properties (and desig- 
nation of National Natural Landmarks) depends, 
in part, on a thorough awareness of the nation's 
biotic diversity. Here again a high-quality 
continental synthesis of knowledge of vascular 
plant species and their geographical and ecological 
distributions would be an invaluable reference 
source for management decisions concerning the 
conservation of our nation's biotic diversity in 
the National Parks and other wild natural areas. 


New York Botanical Garden, Bronx, 
York 10458. 


data bank 

terity ca 

ry and pr 
f ication. 
tion with 

lems rais 
In press. 

Ann. Miss 


1971. Flora Nor 
of systematic bio 

1973. Flora Nor 
sualty? BioScienc 

1975. Recent ad 
actice of biologic 

pp. 11-52 in Bio 
Computers, R. J. 
Press, London and 

Some information 
ed by the internat 
c biology data. P 
nference, Pergamon 

1974. Plant sys 
ouri Bot. Gard. 61 
G. 1971. Flora No 
on system. BioSci 

th America: A 
logy. BioScience 

th America: Aus- 
e 23:215. 

vances in the theo- 
al specimen identi- 
logical Identifica- 
Pankhurst (Ed. ) , 
New York . 
-management prob- 
ional nature of 
roc. 5th Biennial 
Press, Oxford. 

tematics 1947-1972. 


rth America as an 

ence 21:524, 529- 


SHETLER, S. G. 1974. Demythologizing biological 

data banking. Taxon 23:71-100. 
WATSON, L. 1971. Basic taxonomic data; The 

need for organization over presentation and 

accumulation. Taxon 20:131-136. 

9 6 


L. K. Thomas, Jr. 


Theodore Roosevelt Island is 
of the National Capital Region o 
Park Service. Located in Washin 
Potomac River (Figure 1) is is 3 
(88.32 acres) in area (National 
1970:56), approximately 1.1 km ( 
0.5 km (0.3 mile) wide at its wi 
Geological Survey, 1965) . The i 
a core of micaceous schist surro 
(Thomas, 1963:1,7). The highest 
about 13.1 m (43 ft., ibid.). 

one of the parks 
f the National 
gton, D.C. , in the 
5.74 hectares 
Capital Parks, 
0.7 mile) long and 
dest place (U.S. 
sland consists of 
unded by alluvium 

elevation is 

As early as 1792 the upland area was in agri- 
cultural use and the natural vegetation has been 
periodically disturbed from that time until the 
island was acquired by the National Park Service 
in 1932 {ibid., 2, 49). Since acquisition, there 
have also been disturbances to the vegetation from 
developments: highway bridge, monument to Theodore 
Roosevelt, extensive plantings (ibid., 2, 49, 50; 
U.S. National Park Service, 1968:8). 

The natural vegetation of the upland is domi- 
nated by Ulmus americana L. (American elm) , Acer 
negundo L. (boxelder), Morus alba L. (white 
mulberry) , Prunus serotina Ehrh. (black cherry) , 
Fraxinus americana L. (white ash), Liriodendron 
tulipifera L. (tulip-tree), Quercus rubra L. 

(northern red oak) , and Acer saccharinum L. (silver 
maple) in about that order (Thomas, 1974:7). The 
forested alluvial deposits are dominated by Acer 
sacchar inum L. (silver maple) , Fraxinus 
pennsylvanica Marsh, (green ash), and Salix nigra 
Marsh, (black willow) , while the nonforested 
deposits are dominated by a fresh water tidal 
marsh of Peltandra virginica (L. ) Schott & Endl . 

(arrow-arum), Acorus calamus L. (sweetflag), Typha 
angustifolia L. (narrow-leaved cat-tail), and 
Nuphar luteum (L.) Sibth. & Sm. (spatter-dock) 

{ibid . , 7,8). 

Because of their abundance, three exotic plant 
species were selected for ecological study in 1971: 
Hedera helix L. (English ivy), Lonicera japonica 
Thunb. (Japanese honeysuckle) , and iris pseudacorus 
L. (European yellow iris) . The study which was 
terminated in 1974 was carried out to determine 
what native species, if any, were being replaced 
and what factors limit exotic abundance. 

In all six of these habitats square meter 
plots were laid out in several experimental 
designs. In noncontrol plots exotic plant bio- 
mass was removed, dried, and weighed. Invasion 
of native and other species into some of these 
plots was noted. 

Light stations, using Ozalid type light 
meters (Friend, 1961) , were set up at the plots 
and in similar habitats without the exotics. 
Light was determined as a percent of open sun- 
light, therefore, meters were also placed as 
controls in an open field on the island. In 
addition controlled light experiments were run 
in both the English ivy and the Japanese honey- 

Surveys were made of the elm trees on the 
island and their vigor and relationship with the 
two exotic evergreen vines: English ivy and 
Japanese honeysuckle. All down trees on the 
island were surveyed along with their relationship 
with these same exotic vines. 

The topography of the marsh and swamp-marsh 
transition was determined as well as a number 
of soil factors as they related to presence or 
absence of the European yellow iris. 

Altogether there were 190 experiments and 
surveys which were statistically analyzed. 


Light is a strong limiting factor for Japanese 
honeysuckle (Figure 2) . Growth increases with 
light intensity. This evergreen vine is a twiner 
and it overwhelms and kills small trees and 
shrubs, and suppresses reproduction especially 
of the tree species American elm, black cherry, 
and tulip-tree which are among the dominants on 
the island upland. 

Light is not as strong a limiting factor for 
English ivy (Figure 2) , but on the upland no 
factor is stronger. On the flood plain, the 
water table limits growth and distribution of 
the ivy. In both of these habitats the ivy 
suppresses growth of herbs. 

Each of the th 
studied in two hab 
The English ivy wa 
and the flood plai 
studied on the upl 
where subcanopy tr 
in an upland area 
had been removed, 
studied in the ope 
back near the tree 


ree exotic plant species was 
itats in which it was abundant, 
s studied on the island upland 
n. The Japanese honeysuckle was 
and under an intact understory 
ees and shrubs were present and 
where subcanopy trees and shrubs 

The European yellow iris was 
n marsh near the tidal gut and 

line in the marsh-swamp 

On the flood plain 
vine is about the same 
from hurricane Agnes, 
upland herbs exceeds th 
green liana is not limi 
trees; it shades and ki 
overstory trees. Exami 
tions from overstory si 
English ivy shows suppr 
other rings. The Engli 
honeysuckle areas. 

the impact of this exotic 
as the impact of the flood 
The impact of this ivy on 
is. This tendril ever- 
ted to climbing small 
lis out understory and 
nation of log cross sec- 
zed trees covered with 
essed ring growth in the 
sh ivy takes over Japanese 

National Park Service, National Capital Region, 
Washington, D.C. 20242. 

When the canopy is opened by a tree falling, 
vines of both English ivy and Japanese honeysuckle 
are encouraged. A great boost, however, is given 
to English ivy growth when an American elm tree 
is cut (for management of Dutch elm disease) . 


; N \ uKiv>:i, ; m .-^ ■ ... s -__ \ ^ U , £-;P«!i 

-■* 4 , -"" ■ jri 

■ ■, ^.,*,«hin«I9H Doblors *' , 

The Three Sisters 

^ -. B ,. \V Roosevelt g ■ " ... , .'',., «J 

vi,.. r -Fort My<>r ' r >> ' \;" , ,i^ \ -^^^R^__ 

aefliOinaTfloLj Q 


1000 ?000 3000 JOiJO MOO 6000 '000 FEET 


FIGURE 1. Theodore Roosevelt Island and vicinity. 

For the European yellow iris 
time it is inundated comes close 
factor. The longer the inundati 
the growth. By assuming this as 
factor a rise in the river level 
ing season was accurately predic 
marsh increases the microtopogra 
as any raised surface in marshy 
encouraging the iris. The iris 
arum areas. In the swamp-marsh 
sweetflag takes over iris areas, 
ciated with soil hardpan, especi 
transition area. 

the length of 
st to a limiting 
on, the slower 
the limiting 
during the grow- 
ted. Trash in the 
phy and functions 
habitats by 
takes over arrow- 

Iris is asso- 
ally in this 


Once Japanese honeysuckle becomes established, 
its suppression of tree reproduction and killing 
of small trees and shrubs changes not only the 
composition of the forest, but its structure as 
well. With less woody plants the area becomes 
more sunlit, thus Japanese honeysuckle tends to 
create its own best habitat. 

The killing of trees by English ivy appears to 
result both from shading out the tree leaves and 
compressing (in the case of heavy ivy growth) the 








B i om ass 

i n 
g'rams/m 2 








Hedcrq (ivy) 

logY c = log a + b logX 

sig. at 0.025 

r* = 50% 

Lonicerq (honeysuckle) 
Y c = a * bX 

sig. at 0.001 
r2 = 69% 



20 40 60 

Light in % of open 

FIGURE 2. The relationship of biomass 
(English ivy) and l. japonica (Japanes 
vascular tissue to prevent or slow down food 
translocation and water uptake. American elms 
are particularly vulnerable, and this is probably 
due to reduced vigor from Dutch elm disease and 
root compaction from the many trails on the 
island . 

Overall, the consequences of the establishment 
of these two ornamental vines are that the 
forests (except the swamp) are being destroyed 
and succession is being directed toward a vine 
dominated community. The ultimate community will 
be dominated by English ivy and the management 
for the control of Dutch elm disease is an impor- 
tant factor in encouraging this ivy. 

Since the European yellow iris replaces the 
arrow-arum which may be an important wood duck 
(Aix sponsa (L.)) food supply in the nesting sea- 
son (Martin et al . , 1951:447; Stewart and Robbins, 
1958:85), these ducks may be in jeopardy. 
Stewart and Robbins (1958:21) consider this bird 
as one of the primary species of breeding birds 
in the Coastal Plain of Maryland and the District 
of Columbia. Since the island straddles the 
Piedmont and Coastal Plain physiographic pro- 




80 100 % 

sunlight (X) 

to % of open sunlight for H. Helix 
e honeysuckle) . 

vinces (Thomas, 1963: 7, 39), the further possi- 
bility exists of reducing the range of a primary 
species of breeding bird. 

Once the iris gets established on a site, its 
rhizomes ramify the soil and in growth compact 
that soil to create a hardpan. With the drier 
hardpan plus a slightly raised mat of rhizomes, 
more favorable conditions are created for iris 
growth, thus the iris apparently helps to create 
its own best habitat. 

In the swamp-marsh transition area, this iris 
appears to outcompete the willows {salix) for 
mineral soil and thus eliminate the willow stage 
from the normal marsh to swamp succession. It 
thus short-circuits succession and provides a 
seed bed for swamp trees such as green ash. 

At the present time this short-circuit is 
being slowed down because of a rise in the river 
level during the growing season. This makes 
the marsh and adjacent swamp-marsh transition 
wetter for a longer period of time and hence less 
favorable for iris growth. 





.7700 - • 


in m (X) 




.7100-- • 


Ins pttudacorui 
Y c = a ♦ bX 

sig. at 0.001 
r 2 = 47% 

FIGURE 3. The relationship of topography (elevation) 
to biomass in I. pseudacorus (European yellow iris). 

400 600 800 1000 
Biomass in grams/m 2 (Y) 

Another factor operating in the swamp-marsh 
transition is the sweetflag. This native species 
is outcompeting the iris, and in so doing it con- 
firms the r- and K- selection theory. The sweet- 
flag behaves as a K- strategist by putting most 
of its resources into vegetative structure, while 
by comparison the European yellow iris is an r- 
strategist that produces a lot of seed. Accord- 
ing to the theory (see Gadgil and Solbrig, 19 72: 
14, 17-20) the r-strategist (the iris) will be 
most successful in a density-independent mortality 
environment and the K- strategist (the sweetflag) 
will be most successful in a density-dependent 
one. Thus one could predict that in the emergent 
zone of marsh vegetation where both these species 
live, that the K- strategist will do better as the 
zone fills with vegetation and density-dependent 
factors become more important. This is what is 
actually happening in the marsh. 

All three of the exotic plant species studied 
were established before the island became a park 
forty-four years ago. They are all associated 
with disturbed areas, and in every case they are 
altering the native vegetation. 


FRIEND, D. T. C. 1961. A simple method of mea- 
suring integrated light values in the field. 
Ecology 42:577-580. 



GADGIL, M., and 0. T. SOLBRIG. 1972. The concept 
of r- and K- selection: evidence from wild 
flowers and some theoretical considerations. 
The American Naturalist 106 (947) : 14-31 . 

MARTIN, A. C, H. S. ZIM, and A. L. NELSON. 1951. 
American Wildlife and Plants. Dover Publica- 
tions, Inc., New York. 500 pp. 

NATIONAL CAPITAL PARKS. 1970. Reservation List, 
January 1, 1970. United States Department of 
the Interior, National Park Service. 69 pp. 

STEWART, R. E., and C. S. ROBBINS. 1958. Birds 
of Maryland and the District of Columbia. 
Bureau of Sport Fisheries and Wildlife, United 
States Department of the Interior. North 
American Fauna, No. 62. 401 pp. 

THOMAS, L. K., JR. 1963. Geomorphology and Vege- 
tation of Theodore Roosevelt Island. National 
Park Service Scientific Report No. 2. 61 pp. 

. 1974. The impact of Three Exotic 

Plant Species on the Native Vegetation of a 
Potomac Island. Ph.D. Dissertation, Duke Uni- 
versity, Durham, North Carolina. 281 pp. 

ton west quadrangle, 7.5 minute series (topo- 
graphic) . 


Master-plan, Theodore Roosevelt Island, Wash- 
ington, D.C. United States Department of the 
Interior, National Park Service. 52 pp. 



Jill Baron, Christine Dombrowski and Susan Power Bratton 

This report describes the occurrence and status 
of five non-native plant species within the 
Tennessee side of Great Smoky Mountains National 
Park. The research project attempted to determine 
where the plants are located, whether their popu- 
lations are expanding or contracting, and what 
impact, if any, they have on the native flora. 
Management of exotic plants is important for two 
reasons. The first, more vital reason is the pre- 
servation of an irreplaceable native flora. Plants 
that are in a delicate balance with their environ- 
ment may not survive competition with introduced 
species. The second reason is certainly aesthetic. 
Non-native plants growing along the roadsides ob- 
scure the native flora, giving visitors a false 
impression about the structure and composition of 
the plant communities in the park. 

The Resources Management Plan (of 1969) for the 
Great Smoky Mountains proposed that the exotic 
plants in the park be divided into three manage- 
ment groups: 

1. Exotic plants which are displacing present 
native species and control is practical 

2. Exotic plants which are displacing present 
native species and control is not 

3. Exotic plants that remain static or those 
that have been declining in present years 

River Road, Newfound Gap Road, Greenbrier Road, 
the Bypass, the Spur, and Highway 73 to Green- 
brier. The trails were chosen to represent a 
variety of forest types and elevations. Some 
manways and abandoned roads were also covered. 
About three hundred sitings of the five exotics 
in question were recorded. At each site we 
took the following information: 

1. The exact location of the plant (s) , includ- 
ing the coordinates on the 7.5 minute 
topographic maps, and a written description 
of the location 

2. Elevation above sea level 

3 . Slope and aspect - slope was estimated or 
measured using a Brunton pocket transit; 
aspect was measured with a compass 

4 . Site type - the investigators noted 
whether the plants were growing around old 
home sites and fields, along streams or 
rivers, or along roads or trails 

5 . Forest type and understory type - this in- 
cluded the estimated age of the stand 

6. Site disturbance - this included evidence 
of human intervention such as mowing or 
grading, and evidence of recent "natural" 
disturbance such as hog-rooting 

(III-A-33 Resources Management Plan) 

Plants included in Group 1 were kudzu ( Pueraria 
thunberg iana ) , mimosa (Albizzia julbrissin), and 
tree-of-heaven (Ailanthus altissima) . Although 
all three species have been slated for control, 
at present, only an active eradication program 
for kudzu has been initiated. Research has been 
conducted on mimosa leaf wilt at the National Park 
Service Science Center at Bay St. Louis, Mississip- 
pi, but no field tests have been made in the park. 
Japanese honeysuckle (Lonicera japonica ) was in- 
cluded in Group 2. Control was therefore not 
recommended. No specific species were included 
in Group 3 . 

In addition to kudzu, mimosa, tree-of-heaven, 
and Japanese honeysuckle, this study includes the 
Princess tree (Paulonia tomentosa) , which fre- 
quently associates with mimosa, and quickly invades 
large road cuts, particularly those near rivers. 


The field research for this report was carried 
out in the summer of 1975. Two hundred and 
seventy miles of trails and roads on the Tennes- 
see side of the park were observed to determine 
the extent of the invasion of woody non-native 
plants. Roads walked included the Cade's Cove 
Loop Road, Laurel Creek Road, Tremont Road, Little 


National Park Service, Uplands Field Research 
Laboratory, Great Smoky Mountains National Park, 
Gatlinburg, Tennessee. 

Population of the exotic - the investigators 
noted the number of mature plants and 
seedlings present at a site, the presence 
of flowers or fruits, and any indication of 
vegetative spreading and reproduction 

Area of exotic distribution in square 

Affect on neighboring plants - the vines, 
kudzu and honeysuckle could be placed into 
one of three categories according to their 
impact on their closest neighbors. The 
first category was "slight", which indi- 
cated the plant was not growing on other 
species and generally had a scattered 
distribution. Individual plants or shoots 
were small. The second category was 
"moderate", which indicated that the plant 
was growing on other species or making 
solid clumps, excluding other species. 
The third category was "great", which in- 
dicated that the plants were climbing on 
trees or strangling other species. The 
three tree species had little impact other 
than shade on their neighbors, and tended 
to be too scattered to completely exclude 
other species. Therefore, no categories 
of impact were developed for these. 

Nearest neighbors - in order to quantify 
the community type, the investigators 
recorded the names of the ten nearest herb 
species, and the species of each of the 
ten nearest trees (greater than 4 inches in 
diameter at breast height) and of the ten 
nearest shrubs (less than 4 inches DBH) . 



Lonicera Japonica 

In the course of this study, Japanese honey- 
suckle was recorded 86 times. Of these occur- 
rences, 83 percent were in a successional forest 
dominated by Liriodendron tulipifera. The species 
composition of the successional tulip tree for- 
est type (Whittaker 1964, Howe and Bratton 1975) 
varied along a moisture gradient. At the drier 
sites Pinus Virginia , P. strobus , Quercus 
coccinea, Q. rubra and Q. alba were common. 
Trees found at wetter sites were Plantanus 
occ idental is , Juglans nigra, Carpinus caroliniana, 
and Liquidambar styraciflua. Several vines com- 
mon in successional forests were often present as 
nearest neighbors to the honeysuckle. These were 
Rhus radicans , Parthenocissus qu inque folia , Vitus 
spp. , Rubus spp. , and the herbaceous vine 
Amphicarpa bracteata . 

Herbaceous species commonly found growing 
near the honeysuckle along the roadsides in- 
cluded native species such as Fragaria virginiana , 
Potentilla spp., and Polystichum acrostichoides , 
and non-native species such as Plantago major, 
P. lanceolata , Cynodon dactylon , Eluesine indica, 
Tri folium pratens , T. repens , and T. procumbens . 
Around the old homesites Panicum spp., Aster 
spp., Solidago spp., Agrimonia sp. Fragaria 
virginiana, Potentilla spp., Polystichum 
acrostichoides , and the exotic Dioscorea 
batatas were common. Altogether, over a 
third of the herbaceous species associated 
with the honeysuckle were exotics themselves. 

Seventeen percent of the records for the honey- 
suckle were in areas with no mature canopy at all. 
The impact of the honeysuckle on its neighbors 
depended on the presence of a canopy. Of the 
stands occurring under a canopy, one-fourth were 
judged as having heavy impact, half had a moder- 
ate impact, and one-fourth had slight impact on 
its neighbors. This could be due to the compe- 
tition provided by the established grasses. 
Forest floor herbs in contrast, seemed to offer 
less resistance to invading honeysuckle, and in 
some areas are completely excluded by a heavy 
ground cover of the vine. 

The impac 
related to it 
be expected, 
their neighbo 
ally moderate 
suckle clumps 
to those of h 
little impact 
they were spa 
shoots tended 

t of a clump 
s area. The 
tended to hav 
rs. Clumps f 

in their eff 

graded from 
eavy impact. 

were found o 
rse and proba 

to be small 

of honeysuckle was also 
smallest clumps, as might 
e the least effect on 
rom 6 to 30 m^ were usu- 
ect. The largest honey- 
areas of little impact 
The large stands with 
n the roadsides where 
bly frequently cut. The 
and scattered. 

The lowest elevation where 
found was 1,200 feet. The high 
3,360 feet. Most of the sighti 
tween 1,400 and 2,000 feet abov 
plants at the higher elevations 
pale suggesting altitude may be 
tor in the distribution of the 
It is important to note, howeve 
low elevation forest in the par 
A majority of the old homesites 
below 2,200 feet and the forest 
vation are regenerating after f 

honeysuckle was 
est elevation was 
ngs occurred be- 
e sea level. The 
were small and 
a limiting fac- 
species in the park. 
r, that most of the 
k is successional. 

in the park are 
s below this ele- 
arming, logging or 

The impact of the honeysuckle was related to 
elevation. The greatest impact was found in 
stands growing at elevations less than 2,000 feet. 

Above 2,000 feet the density of the stands and 
the impact decreased steadily. Above 3,000 feet 
the stands had almost no effect at all on the 
surrounding plants. 

Honeysuckle was found on slopes ranging from 
degrees to 90 degrees. Over half (63 percent) of 
the sightings were in areas with to 10 degree 
slopes. This appears with site type: 84 of the 
site locations were homesites and roadsides, 
areas which are primarily flat. Twelve of the 
honeysuckle clumps were on slopes greater than 40 
degrees. Of these, two had only slight impact, 
seven had moderate impact, and three had great 
impact on the neighboring plants. 

Seventy-three of the honeysuckle sightings were 
on roadsides, ten on old homesites, and three by 
riversides. Our field time was not divided 
equally between the three different site types, 
since a majority of the time was spent on the 
roads. However, this does not fully account for 
the difference in occurrence by site type. Many 
old homesites investigated had no honeysuckle. A 
homesite with mimosa seedlings would often have 
no honeysuckle and vice versa. Honeysuckle may 
not have been common as an ornamental, or it may 
have been common but unable to survive in the 
shade of the successional forest. The three 
riverside sightings were all between Little River 
Road and Little River. In these areas, the honey- 
suckle was growing up relatively steep banks. No 
honeysuckle was found growing right by the water. 
The presence of Japanese honeysuckle on all of 
the roads investigated suggests that it needs an 
open canopy to become established. Once estab- 
lished, however, it may invade nearby forests. 
Most of the stands of the honeysuckle were con- 
fined to road cuts and forest edges, but several 
dense clumps went back 6 to 8 meters under the 
trees. Two locations on the Little River Road 
above Sugarlands Visitor Center had very dense 
stands that went into the forest at least 15 
meters. The density of these stands decreased 
markedly inside the canopy, presumably due to 
shade . 

Albizzia julibr issin 

In the course of this study, mimosa was 
recorded 87 times. Although 80 of the sightings 
were near tulip tree successional forest, none of 
the mimosa plants were actually growing under 
closed canopy. Seven sightings were in open areas. 
The tree was found growing along the same wide 
moisture gradient as the honeysuckle . Mature in- 
dividuals were observed growing along stream 
banks as well as more xeric areas such as road 
cuts near pine-oak forest. 

The herbaceous species found in association 
with mimosa along the roadsides and around the old 
homesites are largely the same as those associated 
with the Japanese honeysuckle. Trees growing by 
streams rarely had any herbs near them. Some 
sedges were found, as well as Campsis radicans , 
Amphicarpa bracteata and Sol idago spp. 

Only sixteen of the mimosa sightings included 
mature trees or clumps of mature trees; the re- 
maining seventy-one sightings included only 
seedlings. As many as 1,000 seedlings were re- 
corded on one data sheet if the seedlings were 
all in the same vicinity. There were no mature 
mimosa plants around the old homesites which were 


being overgrown by the forest community. At many 
of the other mimosa sites, mowing probably removed 
most of the seedlings the first year. Shade from 
forest trees may also negatively affect seedling 
survival . 

Mimosa was found at a ma 
feet and at a minimum eleva 
It was most common between 
Note that the maximum eleva 
for Japanese honeysuckle wa 
than the maximum elevation 
Mature individuals of mimos 
1,200 and 2,150 feet. The 
found at lower elevations o 
2,400 feet may be about the 
limit of mimosa in the Grea 

ximum elevation of 2,4 20 
tion of 1,160 feet. 
1,200 and 2,000 feet. 
tion presently recorded 
s 1,000 feet higher 
recorded for mimosa . 
a were found between 
species is certainly 
utside the park but 

upper altitudinal 
t Smoky Mountains. 

Most of the mimosa (68 percent) was found 
growing on slopes of to 10 degrees, but the 
slopes of some of the mimosa locations were as 
steep as 90 degrees. 

Mimosa seedlings have very little impact on the 
native flora. They were sparsely spread along 
roadsides and around old homesites. Since mimosa 
seeds are very light, the seed sources for this 
heavy production of seedlings are probably out- 
side the park as well as inside the park boundary, 
seedlings would probably reestablish themselves 
within a year. 

Around homesites, roads and stream banks, 
mimosa was usually found in two sizes: small 
seedlings (6 to 10 cm) or mature trees (up to 10 
meters in height) . The general lack of interme- 
diate growth stages is further evidence that most 
seedlings do not survive their first year, largely 
because of mowing. The lack of intermediate 
stages may also indicate that the mimosa popula- 
tion is stable or declining, outside of the open 
areas around buildings and recent roadcuts. 

Paulownia tomentosa 

Of the eight individuals of princess tree 
found, five were growing on or at the base of 
cliffs. The other three were growing on open 
ground close to streams. Four of the sites had 
slopes of less than 10 degrees and four of the 
sites had slopes of 80 to 90 degrees. The tree 
was found at a maximum elevation of 1,716 feet 
and a minimum elevation of 1,160 feet. Six trees 
were growing in isolated positions, away from any 
other trees. The other two, which were growing 
by streams, had Platanus occidental i s , Liriodendron 
tulipfera , and Betula lenta as nearest neighbors. 
Princess tree is not common in the park and seems 
to grow only in areas where it has little 

Ailanthus altissima 

Only one stand of tree-of-heaven was found in 
the sample area. There were more than twenty 
stems and many seedlings. This group is possibly 
a single clone, originating from one root. One 
tree was in fruit. Neighboring trees were Prunus 
sp., Malus sp., Plantanus occidental is , and 
Liriodendron tulipifera. Shrubs present were 
Aralia spinosa, Robinia pseudo-acac ' a , and 
Liquidambar styrifolia. The herbs were typical of 
roadside successional flora: Ambrosia sp., 
Albizzia julibrissin (present as a seedling), 
Fragaria virginiana, Panicum Sp . , Plantago major, 
P. lanceolata , Solanum carol inese and Viola sp . 

Pueraria thunberg iana 

There is an active management program to elimi- 
nate kudzu from the Great Smoky Mountains National 
Park. No kudzu was found in the course of this 
study . 


Looking at the distributions of exotic plants 
for the Tennessee side of Great Smoky Mountains 
National Park, several patterns of importance to 
management appear. These include: 

1 . Exotic vascular plants are generally not 
found in virgin forests in the Smokies. 
Unlike the wild boar (Sus scrofa) , which 
has invaded forests where there was a mini- 
mum of previous human interference, the 
exotic plants now present in the park are 
dependent on human activity, such as mowing 
or grading, to provide a suitable habitat 
for their establishment. The native vege- 
tation seems to be able to out-compete 

the exotics, as long as the native forest 
communities are not subject to unnatural 
disturbance . 

2. The woody exotics appear to be confined in 
their altitudinal distribution. Even though 
the roads in the park extend above 6,000 
feet, the woody exotics drop out at 3,400 
feet and below. The higher elevations may 
not be subject to invasion by non-native 
woody exotic species, or any invasion may 

be slower and easier to control. There are, 
of course, vigorous clumps of exotic herbs 
present along the roads, even at the 
highest elevations. 

The non-native woody species 
confined to low elevations, b 
confined to one general type 
forest, the tulip tree type a 
tions . This forest type itse 
an artifact of human activity 
the Japanese honeysuckle, the 
do not grow well under the fo 
tend to be along roads, and i 
around old buildings. This f 
the distribution of these spe 
park . 

are not only 
ut are also 
of successional 
nd its varia- 
lf is largely 
Aside from 
woody exotics 
rest canopy and 
n the openings 
urther restricts 
cies in the 

The populations of these species in the park 
are not very large. If one were to consider 
only the saplings and the mature individuals 
of the three tree species surveyed, there 
were barely 20 sightings over 270 miles of 
roads and trails (and the area sampled in- 
cluded some of the best possible habitat 
available in the park) . 

The age structure of the populations does 
not indicate an impending rapid expansion 
of the species. The lack of large number 
of saplings of the three tree species im- 
plies that few seedlings are surviving, 
and that in the next few years there will 
be few additional mature plants to replace 
those already in the park. 

The species included in this survey have 
been present in the area of the park for 
quite a number of years, but have not under- 
gone a real population explosion (unlike 
the wild boar or the chestnut blight) . A 
vine like Japanese honeysuckle probably 
spread quickly after the abandonment of the 


agricultural land in the park, but the same 
species may now actually be losing area to 
the new growth of forest. Unfortunately, 
since no datum has been kept on the status 
of exotic species in the park, one cannot 
determine at present if the populations of 
these species are increasing or decreasing. 

7. With the exception of the Japanese honey- 
suckle, these species have very little 
impact on the native flora. The popula- 
tions are too small and scattered and are 
too restricted to disturbed sites to have 
much effect. 

8 . These woody exotics tend to be along the 
roads and near park buildings, where they 
are highly visible to visitors. In fact, 
the areas which have the heaviest visitor 
use have the greatest numbers of exotics. 


Prevention of further problems with exotic plants: 

Although there 
vascular plants wi 
communities in the 
cies make problems 
First, and foremos 
forest communities 
possible. This is 
case of virgin for 
should be minimal. 
as reservoirs for 
the main invasion 
park. Large road 
for species like m 

is no way to insure that exotic 
11 not invade the native plant 
park, certain management poli- 
with exotics less likely, 
t, disturbance of the native 
should be avoided as much as 
especially important in the 
est. Cutting and grading 

Cleared and mowed areas serve 
vascular exotics. Roads are 
routes into the center of the 
cuts provide excellent habitat 
imosa and princess tree. 

Secondly, grazing damage should be kept to a 
minimum. Overgrazing by deer, wild hogs or other 
species may give exotic species a competitive ad- 
vantage over the native inhabitants of a site. 

Thirdly, the Park Service should be careful 
about introducing exotic species not already pre- 
sent in the park. When possible, native species 
should be planted around the visitor centers and 
residence areas. Care should be exercised when 
reseeding disturbed areas. 

Management for the non-native species now present 
in the park: 

Before outlining a management plan for exotic 
plants, it is necessary to realize that certain 
park programs tend to encourage an exotic flora. 
The park will never be completely free of exotic 
vascular plants, as long as there are roads and 
lawns in the park. Most of the non-native herbs 
established at these sites are not a threat to na- 
tive plant communities, however, and the non- 
native herbs are largely confined to the disturbed 
habitats. Several exotic species are relics of 
earlier agricultural practices and tend to occur 
in areas now managed as historical sites rather 
than as natural areas. Cades Cove, for instance, 
probably has a greater biomass of exotic plants 
than any area of comparable size in the park. 
Many of the forage plants on which both the deer 
and the cattle subsist are exotics. 

Because of the impossibility of eliminating 
vascular exotics from the park (and it remains to 
be seen if this is even desirable) , the investi- 
gators suggest dividing the park into zones for 
the purpose of managing exotic vascular plants . 
The zones are: 

ZONE 1: The exotic-free area - this includes the 
virgin forest areas in the park, forest and heath 
balds with little previous human disturbance, and 
the high elevation forest, both logged and unlogged, 
which have no exotics at the present time. Most 
of these areas support few, if any, exotics now 
and are the areas where invasion by exotics is 
least desirable. At the moment no active manage- 
ment program is needed in these areas, but any 
stands of woody exotics which are discovered should 
be removed at once. Botanists visiting these sites 
will hopefully report any unusual plants to park 
personnel . 

ZONE 2: The buffer area - this includes sites 
which adjoin areas in Zone 1 and roads and disturbed 
areas which pass through areas in Zone 1 . Ab- 
solute exclusion of woody exotics should also be 
practiced here. At present only the Newfound Gap 
Road needs to be watched. 

ZONE 3: The non-management area - this includes 
low elevation second growth forests, old fields, 
old home sites and other successional areas toward 
the outside edge of the park and generally below 
3,000 feet in elevation. These areas have large 
populations of exotics, but are largely inaccessi- 
ble to control. No management is recommended 
unless future study indicates expansion of the 
exotic populations in these areas and an increase 
in the impact on the native flora. Present data 
indicate that normal successional processes will 
tend to slowly exclude many of the non-native spe- 
cies and reduce their populations. 

ZONE 4: The managed area - this zone includes 
roadsides, visitor use areas, residence areas, and 
other places subject to continual human distur- 
bance. These are the sites within the park which 
have the greatest populations of exotics, and are 
also the areas most visible to the public . 

The control of some of the woody exotics could 
well be conducted as a program incidental to 
normal maintenance. In many of the areas with 
concentrations of exotic plants, the Park Service 
is cutting or mowing. If the maintenance crews 
could identify small saplings of the species in 
question they could pull them out by the roots. 
In general, with mimosa and the other woody exotics, 
removing seedlings is probably more trouble than it 
is worth. Saplings are the best age class to 
attack. Root-sprouts can be controlled by recut- 
ting at close time intervals. Honeysuckle might be 
cut off at the base. 


BRENDEN, E. V. 1960. 
of honeysuckle and 
Mtg. Weed Conf. 18 

BROCKMAN, C. F. 1968 
Golden Press. New 

FERNALD, M. C. 1970. 
9th Ed. D. Van Nos 
1632 pp. 

HOWE, T. D. and S. P. 
rooting activity o 
Castanea . 41:256- 

the administrative 
parks and national 
significance (Natu 
Govt. Printing Off 

Progress report on control 
kudzu. Proc. 13th Annual 

Trees of North America. 
York. 280 pp. 
Gray's manual of botany, 
trand Co., Cincinnati. 

Bratton. 1976. Winter 

f the European wild boar. 


1970. Compilation of 
policies for the national 
monuments of scientific 

ral Area Category). U.S. 

ice. Washington, D.C. 147 pp 


NATIONAL PARK SERVICE. 1974. Final environmental SYMONDS, G. W. D. 1958. The tree identification 
statement natural resources management plan. book. William Morrow and Co., Inc. New York. 

Hawaii Volcanoes National Park. Offset 143 pp. unnumbered. 

guide to wildf lowers of northeastern and north- 1974. Seeds of Woody plants in the United 
central North America. Houghton Mifflin Co. States. U.S. Dept . Agric, Agric. Handb. 450. 

Boston. 420 pp. U.S. Govt Printing Office, Washington, D.C. 

RADFORD, A. E., H. E. AHLES, and C. R. Bell. 1968. 883 pp. 

Manual of the vascular flora of the Carolinas. WHITTAKER, R. H. 1956. Vegetation of the Great 
Univ. of N. C. Press, Chapel Hill. 1183 pp. Smoky Mountains. Ecol. Mong . 26:1-80. 

SYMONDS, G. W. D. 1963. The shrub identification 
book. M. Barrows and Co., New York. unnum- 
bered . 



D. J. Frederick 2 , L. Rakestraw 3 , C. R. Eder2, R. A. Van Dyke 2 , 
B. J. Griewe 2 , and M. A. Anderson 2 


Pictured Rocks and Apostle Islands National 
Lakeshore are two of the latest additions to the 
National Park System. Pictured Rocks, established 
in 1966 with an authorized area of 66,500 acres 
(26.912 ha), stretches 35 miles (56.32 km) along 
the southern shore of Lake Superior between the 
towns of Munising on the west and Grand Marais on 
the east of Alger County, Michigan. The Apostle 
Islands Lakeshore was created in 1970 and encompasses 
42,011 acres (17,002 ha) consisting of 20 islands 
and a section of mainland on the Bayfield peninsula 
in Ashland and Bayfield counties, Wisconsin. 

Both lakeshores offer an unsurpassed variety of 
natural scenic beauty. Pictured Rocks boasts 
multicolored sandstone cliffs, broad sand beaches, 
sand bars and dune, and numerous waterfalls and 
streams. The Apostle Islands archipelago, the north- 
ernmost land area in Wisconsin offers densely for- 
ested islands ranging in size from 3 to 10,000 
acres (7.4 to 2,471 ha) sand beaches, rocky shore- 
lines with sculpured arches and a unique early 

A variety of soils are found ranging from those 
of sandstone origin on both areas with varying mix- 
tures of organic material and glacial drift to the 
famous red clays of the Apostles. Several outwash 
features occur at Pictured Rocks including the 
Kingston Plains and the Grand Sable Dunes. 

The present vegetation of both lakeshores is quite 
varied, the composition depending on topography, 
soil conditions, land use and fire history. Bet- 
ter-drained upland areas generally support Northern 
hardwoods with sugar maple (Acer saccharum Marsh.), 
yellow birch (Betula alleghaniensis Britton) and 
basswood (Tilia americana L.). Beech (Fagus grandi- 
folia Ehrh.) is a common associate on such mesic 
areas at Pictured Rocks, while it is completely 
absent on the Apostle Islands. Northern red oak 
(Quercus rubra L. ) is often an associate on the 
Apostle Islands, while it is less common at Pic- 
tured Rocks. On both areas, the poorer drained 
upland soils support hemlock (Tsuga canadensis (L.) 
Carr.), red maple (Acer rubrum L.) and elm (uimus 
americana L. ) . The sand and outwash soils support 
white pine (Pinus strobus L.), red pine (Pinus 
resinosa Ait.) and jack pine (Pinus banksiana Lamb.). 
The lowland areas support stands of northern white 
cedar (Thuja occidentalis L.), tamarack (Larix 
laricina Du Roi K. Koch), balsam fir (Abies bal- 
samea (L. ) Mill.) and black spruce (Picea mariana 
Mill B.S.P. ) . 

Vegetation surveys have been made of Pictured 
Rocks (Read 1975) . However, recent information is 
lacking for most of the Apostle Islands. Beals 
and Cottam (1960) made the first extensive survey, 
and only recently individual islands have been sam- 
pled (Larson 1975; Northland College 1975; Stadnyk 
et al. 1974) . 

Supported by a National Park Service contract 
to Michigan Technological University. 
Department of Forestry, Michigan Technological 

University, Houghton 49931. 

Department of Social Sciences, Michigan Techno- 
ogical University, Houghton. 

The forests of 
gone significant 
lization. Loggin 
agriculture plus 
land, have all co 
Thousands of acre 
magnificent stand 
hemlock in the ea 
on the Kingston P 
only a scattering 
being the result 
peated wild fires 
originally suppor 
of sugar maple, y 
repeatedly high-g 
support scattered 
origin hardwoods 

these two lakeshores have under- 
changes with the advance of civi- 
g, quarrying, land clearing for 
repeated fires over much of the 
ntributed to these changes, 
s of land originally supporting 
s of white pine, red pine and 
stern portion of Pictured Rocks 
lains is now open grassland with 

of tree cover. The present state 
of exploitative logging and re- 
Islands in the Apostles which 
ted dense, high quality stands 
ellow birch and red oak have been 
raded, often burned and now 

unmerchantable culls, sprout 
and poor quality aspen. 

The early history of man's incursions into these 
two areas began with French explorers in the early 
1600's and fur trading in 1659 (Ross 1960). Later, 
missionaries and more fur traders and the outpost 
of LaPoint was established on Madeline Island in 
the Apostles. During the early part of the 19th 
century, the fur trade declined and commercial 
fishing activity increased. By the mid-19th 
century, explorers like Douglass Houghton, Lewis 
Cass, Henry Schoolcraft, C. S. Jackson and Joseph 
Norwood had arrived. They made observations and 
wrote descriptions of portions of both areas. 
Surveyors arrived during this time to establish 
the township and section lines and to describe the 
physical features of the land. 

The Original Survey and Surveyors 

Several methods can be used to help reconstruct 
original forest. These include inference based 
on existing present-day forests, examiniation of 
old growth remnants, historical accounts and rec- 
ords of early land surveyors (Daubenmire 1968). 
Original surveyor's records have been used by many 
researchers to help construct vegetation maps 
(Kenoyer 1933; Elliott 1953; Potzger et al . 1956; 
Lindsey et al. 1965) . They have been the most com- 
monly used method of reconstructing early forests 
(Bromley 1935) . They possess the special advan- 
tage of having been written on the spot accordinq 
to a previously determined plan and constitute a 
definite sample of the vegetation (Bourdo 1956) . 

The township lines of Pictures Rocks were est- 
ablished by William Burt in 1841. Burt was Michi- 
gan's master surveyor and inventor of the solar 
compass (Chope 1976; Stewart 1935). He is recog- 
nized for accuracy and completeness of his surveys 
and descriptions (Bourdo 1954; Laapala 1974). He 
was particularly concerned with his crew's accuracy 
in addition to his own. Once he found fraud by 
one of his crews, and resurveyed alone seven town- 
ships on his own time and expense (Bourdo 1954) . 

Fellow surveyors of the time were equally im- 
pressed with Burt. William Ives wrote of Burt's 
knowledge of surveying, completeness of his work 
and dedication to his job while often working under 
very difficult circumstances (Ives 1850). 

William Ives and George Adair established the 
interior section lines at Pictured Rocks between 
1850 and 1855. Ives had previously worked with 
Burt and displayed much the same skill in surveying 


and ability to write detailed descriptions of his 
observations. Like Burt, Ives was an exceedingly 
conscientious field man and often wrote more than 
was required. Ives has been described in Hubbard 
(1898) by Lane "... I cannot let Ives' name go 
by without special note of the fine character of the 
work he did. Working far from civilization, through 
an extremely rough and densely wooded country . . . 
he turned out work far surpassing that customary on 
the linear survey . . . . " In addition to his survey 
records, Ives kept a personal diary which has given 
us a closer look at his personality. Both men did 
work of the highest quality and much of what they 
wrote is valuable historical information. 

vegetation and their notes reflect this. 

While establis 
quired to record 
line intersection 
marshes and the i 
were required to 
features and thei 
descriptions for 
and detailed reco 
sions at the Apos 
of streams, lakes 
were found on the 
Stockton Island. 

hing lines, surveyors were re- 
major changes in topography, soils, 
s with lakes, streams, and 
ncidence of windfalls. They also 
estimate the size of these latter 
r distance to the corner. Again, 
Pictured Rocks show a very complete 
rd while there were obvious omis- 
tle Islands. Several instances 
and marshes not being recorded 
western half of Outer Island and 

George Adair was also noted for good quality 
field work at Pictured Rocks. Apparently, his close 
association with Ives resulted in his conscientious 
recording of field data and detailed descriptions 
of the area's physical features. 

The township and section lines of the lands with- 
in the Apostle Islands Lakeshore were surveyed 
during 1852 and 1857 by three separate surveyors; 
J. Allen Barber, Elward L. Baker and Elisha L. 
Norris. Very little written information exists on 
these men compared to that of the Pictured Rocks 
surveyors. An insight into their personalities must 
be made through an analysis of their written survey 
descriptions and comparing them with the work of 
men like Burt and Ives. 

Land subdivision 
dardized but during 
and the Apostle Isla 
tions for surveyors 
instructions for 185 
tically all surveys 
system in 1910 (Dodd 
between instructions 
18 5 5 were in the amo 
the manner in which 
structions specified 
corded both at towns 
corners on township 
of each bearing tree 
the corner recorded, 
the corner, and scri 

techniques today are well stan- 
the period when Pictured Rocks 
nds were surveyed, the instruc- 
changed (Stewart 1935) . The 
5 were the guide book for prac- 
until the end of the contract 
s 1943) . The major differences 

of the 1840' s and those of 
unt of details required and 
they were recorded. All in- 

that four bearing trees be re- 
hop corners and at section 
and range lines. The diameter 

was estimated and distance to 
Each tree was blazed facing 
bed with the letters "B.T". 

Early instructions specified that bearing trees 
at quarter corners were to be located in adjacent 
quadrants, while later instructions specified they 
be in opposite quadrants. 

At Pictured Rocks, Ives and Adair recorded four 
bearing trees at interior section corners in addi- 
tion to those on township and range lines. At 
Apostle Islands only two trees were recorded at 
interior corners. 

The instructions of 1843 and 1850 stipulated 
that at least one or two trees directly intersected 
by a survey line be designated as line trees and be 
recorded as to species, estimated diameter, and 
exact distance between every two corners. Instruc- 
tions of other years stipulated that all trees which 
the line intersected be recorded. Line trees were 
double blazed, one blaze directly above the other. 
The Apostle Island surveyors failed to record any 
line trees. At Pictured Rocks usually two to four 
were recorded on each section line. 

The recording of other details by the surveyors 
is quite interesting and has been useful in our 
delineation of the presettlement vegetation recon- 
struction. Major changes in cover types were re- 
orded and chainages given. Also, understory shrubs 
and herbs were sometimes recorded. At each section 
corner, tree species were recorded in order of their 
prevalence on the preceding section line. Very com- 
plete information of this type was recorded for 
Pictured Rocks while at the Apostle Islands, the 
surveyors evidently had limited knowledge of forest 

When the survey of a township was completed, the 
surveyor was required to draw a plan and write an 
overall description. The township descriptions 
for Pictured Rocks were detailed and often contain- 
ed several paragraphs. In addition to physical 
descriptive information, personal accounts of the 
men, their thoughts and experiences were included. 
Conversely, township description for the Apostle 
Islands were brief, usually not more than a few 
sentences. Therefore, on the Apostle Islands, we 
placed more emphasis on bearing trees in our work 
of reconstructing the original forests. 

A number of variables will influence the validity 
and usefulness of any given survey for delineating 
vegetation types. Bourdo (1956) has described in 
detail variables contributing to errors in type 
map construction from original notes and how one 
could determine any bias.^ 

Even though disadvantages exist, many of the ori- 
ginal surveyors produced accurate and detailed work 
and it constitutes by far the best record we have 
of original vegetation types. Other methods such 
as memories of persons familiar with the area are 
frequently inaccurate due to lack of scientific 
training. Only records written down shortly after 
observation can be trusted and in some instances 
these must be interpreted in terms of the back- 
ground and personality of the writer at the time 
(Duabenmire 1968) . Although true randomness is 
not an attribute of the General Land Office Survey, 
the survey records do constitute a sample of the 
vegetation recorded on the spot according to a pre- 
determined plan. They constitute a definite sample 
of the vegetation and thus are usable for quanti- 
tative as well as qualitative analysis (Bourdo 1956). 


Data were transcribed from the original survey 
notes of the Pictured Rocks and Apostle Islands 
which were available on microfilm. U.S. Geological 
Survey maps and county soil survey maps were used 
in conjunction with the survey notes for final 
forest vegetation typing. 

From the survey notes, the following information 
was recorded: 

Whether bearing and line trees were chosen by 
the original surveyors without regard to species, 
conditions, size or other factors and represented 
a random sample has been considered by other re- 
seachers. Tree attributes such as decay resistance 
of the wood (witness tree durability) , bark smooth- 
ness (scribing ease) , species commonness (corner re- 
location made easier) , or wood value (probability 
of being cut in the future) may introduce bias in 
selecting trees for witnessing corners. Obviously 
selection of bearing trees was not strictly random. 
The major concern is not whether bias was present, 
but whether it would negate the use of survey data 
for type map construction. There presumably is less 
problem with line trees for their recording depend- 
ed strictly on location. However, not all trees 
falling "on line" were necessarily recorded. 


1. The species and diameter of all bearing 
trees and their direction and distance from all 
section corners. 

2. The species, diameter and location of all 
line trees. 

3. The intersection of all lines with natural 
or other noted features. 

4. The species prevalence on each section line 
and the intersection of any forest type boundaries. 

5. Notes on animal life and evidence of human 
presence or settlement. 

All information was transcribed to blank township 
sheets at a scale of two inches per mile. 

A one-way analysis of variance was used to de- 
termine if species or diameter bias was evident for 
section and quarter corner trees. Each tree wit- 
nessing a section or quarter corner constituted one 
observation from the population. The extent of bias 
between species and within diameter classes was 
tested using the Neuman-Keuls method, a variation 
of the Q-test (Snedcor 1956) . 

Importance values (Mottam and Curtis 1956) were 
computed for all species recorded by the surveyors 
by summing relative density, relative abundance 
and relative frequency. This was done separately 
for both lakeshore areas. Each section and quarter 
corner was considered a sample point for determining 
relative frequency. 

For map placement of forest type line boundaries, 
the species and number of occurrences of each bear- 
ing and line tree for each section line was first 
listed. For the Apostle Islands, only bearing trees 
were used because line trees were not recorded 
during the survey. The product of the importance 
value and the number of accurrences of each species 
per section line was calculated. This total was 
then compared to the list of species in their order 
of prevalence originally recorded by the surveyors 
for each section line. With few exceptions the 
ordering of species on these two lists was similar. 
Adjacent section lines were similarly inspected 
and the degree of contagion of bearing and line 
trees noted. Each type was named on the basis of 
the dominant species. Dominance was based on the 
surveyor's species prevalence list for each section 
line and the product of bearing and line tree occur- 
rences by species and their respective importance 
values. U.S. Geological Survey maps were then su- 
perimposed on the data sheets for elucidating topo- 
graphy, and detailed soils maps consulted before 
final type line placement was made. Where possible, 
classification and naming of forest types followed 
the major cover type classification of the Society 
of American Foresters (SAF 1940) . At Pictured 
Rocks, seven types were classified according to SAF. 
Due to the complexity and alternation of dominant 
species on the mesic sites, four new types were 

On the Apostle Islands, twelve types were delin- 
eated, ten of which followed the SAF, while two 
were created. 

A method was attempted at c 
the computer all information u 
line designations. This was d 
for Pictured Rocks. We are at 
a plotted type map by inputing 
and calculating importance val 
this with soils and topographi 
hope such a computer generated 
able in accuracy and detail wi 

oding and storing in 
sed anually in type 
one on a trial basis 
tempting to produce 

bearing tree data 
ues and integrating 
c information. We 

map will be compar- 
th the hand drafted 

determined from existing timber type maps, aerial 
photographs and ground reconnaissance. For Pic- 
tured Rocks some timber maps were available for 
former Cleveland-Cliffs Iron Company lands now with- 
in the Lakeshore boundary. For the Apostle Islands, 
type map information was available from published 
sources (Larson 1975; Northland College 1975; 
Stadnyk et al. 1974) , state and federal agencies. 
Vegetation data collected by Northland College dur- 
ing 1974-75 were also used to construct type maps of 
several islands. Based upon this information, a 
generalized comparison was made with the original 


From the original surveyor's data and descrip- 
tions of the Pictured Rocks and Apostle Islands 
areas, vegetation type maps of considerable detail 
were drafted using bearing trees, soils and topo- 
graphic information. Included on these maps were 
locations where the surveyors recorded animal sight- 
ings, windthrows, natural features and evidence of 
human presence. 

Results of 
quarter corner 
veyors for sel 
After close in 
ed and not con 
example, small 
intolerant or 
as witness tre 
Other studies 
for medium siz 
large, old and 
where possible 

the statistical tests on section and 

trees showed some bias by the sur- 
ecting certain species and diameters, 
spection, the bias found was expect- 
sidered critical in this work. For 

diameter trees (less than 6 inches), 
short-lived species were not chosen 
es when other species were available, 
have shown preference by surveyors 
e trees (Bourdo 1956) . Exceptionally 
defective trees were also avoided 

The present vegetation cover of both areas was 


The attempt at producing a computer generated 
type map is incomplete. However, the three sets of 
data including: bearing trees, soils and topo- 
graphic information were successfully stored in the 
computer in a form usable for assigning types. A 
program was also developed that assigns cover types 
based on importance values to each section and 
quarter corner. A program to incorporate soils 
and topographic information with the bearing tree 
data is still to be completed. However, a program 
which produces a listing of all three data sets 
for each section and quarter corner including im- 
portance value cover types is complete and can be 
used to simplify hand calculations. 

Large scale exploitation and disturbance of the 
original forests of both areas began in the 1880 's 
with the logging of the white pine. In the following 
paper, Dr. Rakestraw will discuss the land use his- 
tory of the Apostle Islands in detail. Therefore, 
I will concentrate on the Pictured Rocks area and 
identify the land use practices which have had the 
greatest influence on the complexion of the present 
forest vegetation. 

The relative effect of various land use prac- 
tices at Pictured Rocks has depended on several 
factors, one of the most important being the site. 

Little compositional change has occurred on the 
bottomland sites. Much of the spruce, fir, cedar 
and tamarack has been cut only lightly. These areas 
have generated very well. 

The mesic forests have undergone moderate compo- 
sitional change and reduction in average tree size. 
High value species such as sugar maple and yellow 
birch have been cut heavily. Hemlock has also been 
cut heavily and has suffered from subsequent fires 
and post logging decadence. Records of cuttings 
on former Cleveland-Cliffs Iron Company lands in the 
eastern portion of the Lakeshore show heavy selec- 
tive cutting beginning in the early 1940 's. Approx- 
imately 50% of the volumes were removed in the 
form of merchantable logs and chemical wood. These 

cuttings continued on a 10-20 year interval until 
the Park Service acquired the lands. More clear- 
cut logging was done in the western sections for 
charcoal wood (Richard Ewalt, pers. comm. ) . 

A major compositional change is presently occur- 
ring in the wetter mesic forest areas with the elim- 
ination of the American elm by Dutch elm disease 
(Ceratocystis ulmi (Buism. ) C. Moreau) . 

The xeric forest composed of white pine, white 
pine-red pine and white pine-hemlock types has un- 
dergone the greatest change since the original sur- 
vey. Virtually all the forests of these types has 
been destroyed. Clearcut logging followed by re- 
peated fires not only eliminated the few remaining 
uncut trees and logging slash but also destroyed 
the thin organic layer, exposing the bare sandy 
soil. Today the area supports mainly grasses, 
braken fern {Pteridium aquilinum (L.) Kuhn.) and 
a few scattered clumps of aspen, (Populus tremu- 
loides Michx.) white birch {Betula papyri f era 
Marsh.) and small pine plantings. 


By comparing and contrasting the surveyors and 
the data and descriptions they recorded for these 
two areas, several factors become apparent. Con- 
siderable difference exists between the surveyors 
of these two areas and the detail of their work. 
This directly affected the use of these data for 
reconstructing the original forest vegetation. Al- 
though selection of bearing trees by the surveyors 
was not truly random, the survey records did con- 
stitute a vegetation sample recorded according to 
a previously determined plan. 

The value of type maps of any area using ori- 
ginal survey notes are contingent upon the preced- 
ing factors plus the instructions to the survey 
crews at the time the survey was done. If research- 
ers acquaint themselves with this information and 
establish the validity of the data they use, vege- 
tation maps can be produced that are accurate and 
have several practical uses. 

One of the primary values of such maps is their 
use as an ecological data base for planning and 
management. By determining the original forest 
vegetation, a better understanding of the potentials 
of the area can be realized. Also by comparing the 
original forests with present conditions, a better 
understanding can be gained of the influence of 
natural and man-related disturbances and the local 
processes of succession. For example, the use of 
fire as a management tool might be adopted more 
readily after understanding its natural place in 
succession on certain sites. 

Finally, for these two areas our original vege- 
tation map should be a valuable addition to the 
interpretive program. Vegetation changes over the 
last 125 years have often been dramatic. To ac- 
quaint the public with this history not only makes 
them better informed but also allows them to ap- 
preciate and understand management decisions made 
by the Park Service on these public lands. 


BEALS, E. W., and G. COTTAM. 1960. The forest vege- 
tation of the Apostle Islands, Wisconsin. Ecol. 

BOURDO, E. A. Jr. 1954. A validation of methods used 
in analyzing original forest cover. Ph.D. 
Diss. Univ. of Michigan, Ann Arbor, Michigan. 
207 pp. 

BOURDO, E. A. Jr. 1956. A review of the general land 
office survey and of its use in quantitative 
studies of former forests. Ecol. 37:754-768. 

BROMLEY, S. W. 1935. The original forest types of 
southern New England. Ecol. Mono. 5:61-89. 

CHOPP, I. E. 1976. William Austin Burt — Surveyor 
and inventor (1792-1858). Surveying and Map- 
ping 36:249-253. 

COTTAM, G, and J. T. CURTIS. 1956. The use of dis- 
tance measurements in phytosociological samp- 
ling. Ecol. 37:351-360. 

DAUBENMIRE, R. 1968. Plant Communities: A text- 
book of synecology. Harper and Row, New York. 
300 pp. 

DODDS, J. S. 1943. Original instructions governing 
public land surveys (1815-1955). Powers Press, 
Ames, Iowa. 215 pp. 

ELLIOTT, J. C. 1953. Composition of upland second 
growth hardwood stands in the tension zone of 
Michigan as affected by climate and man. Ecol. 
Mono. 23:271-288. 

HUBBARD, L. L. (ed.). 1898. Alfred G. Lane, "Isle 
Royale" in: Geological Survey of Michigan. 
Vol. VI, p. 4. 

IVES, W. 1850. Diary of William Ives. State His- 
torical Collections, Ann Arbor, Michigan. 

KENOYER, L. A. 1933. Forest distribution in south 
western Michigan as interpreted from the origin- 
al land survey (1826-1832). Mich. Acad. Sci. , 
Arts, and Letters, Papers. 19:107-111. 

LARSEN, W. C. 1975. Vegetation ecology of Bear 
Island — Apostle Islands National Lakeshore. 
Masters Thesis, Michigan Technological Univer- 
sity, Houghton, Michigan, 91 pp. 

LAPPALA, D. D. 1974. Retracement and evidence of 
public land surveys. U.S. Dep. Agr. For. Serv. 
Rpt. 52 pp. 

1965. Soil relations and distribution map of the 
vegetation of presettlement Indiana. Bot. Gaz. 

NORTHLAND COLLEGE. 1975. Vegetation surveys of the 
Apostle Islands. Unpublished data held by Sigurd 
Olson Instit., Northland Coll., Ashland, Wise. 

1956. The forest primeval of Indiana as record- 
ed in the original U.S. land surveys and an 
evaluation of previous interpretations of 
Indiana vegetation. Butler Univ. Bot. Stud. 

READ, R. H. 1975. Vascular plants of Pictured Rocks 
National Lakeshore, Alger County, Michigan. 
Mich. Bot. 14:1-43. 

ROSS, H. H. 1960. LaPoint — Village outpost. Edwards 
Bros. Inc., Ann Arbor, Michigan. 200 pp. 

TYPES. 1940. Forest cover types of the eastern 
United States. Soc. Amer. For. Washington, D.C. 
39 pp. 

SNEDECOR, G. W. 1956. Statistical methods. (5th ed.) 
Iowa State Univ. Press, Ames, Iowa. 534 pp. 

STADNYK, L., R. L. VERCH, and B. A. GOETZ . 1974. 
Stockton Island survey: An ecological survey 
and environmental study of Stockton Island, 
Apostle Islands National Lakeshore, Northland 
Coll. Rpt. , 71 pp. 

STEWART, L. O. 1935. Public land surveys: history, 
instructions, methods. Meyers Printing Co., 
Minneapolis, Mn. 202 pp. 




2 3 3 

L. Rakestraw, D. J. Frederick, C. R. Eder , 

R. A. Van Dyke, and B. J. Griewe" 


In 1930, Harlan Kelsey, a National Park Service 
Collaborator, made a report on the Apostle Islands. 
In it he wrote: 

What must have been once a far more striking 
and characteristic landscape of dark coniferous 
forest has been obliterated by the ax followed 
by fire. 

This unfortunately is the most universal 
asoect of the islands themselves as viewed from 
the water — the virgin forest having been ruth- 
lessly exterminated, mostly within the last 5- 
20 years. It will probably take 50 to 100 years 
to reproduce, even in a measure, those wonder- 
ful oak and evergreen growths and to restore the 
original charm and beauty, of this mature pri- 
meveal forest. The ecological conditions have 
been so violently disturbed that probably never 
can they be more than remotely reproduced. .. the 
hand of man has mercilessly and in a measure, 
irrevokably destroyed their virgin beauty. . . 
(Kelsey 1931) . 

What were the patterns of land use and abuse 
that called forth Kelsey 's comment 4 5 years ago? 
To answer this, we must examine the land use of 
the area from an historical point of view. 

The twenty-two Apostle Islands are situated north 
of Ashland, Wisconsin. The number of islands has 
changed over the years as Sand and Rock Islands have 
been destroyed by wind and wave erosion. At the 
time Captain H. W. Bayfield made his geological 
survey of the lakes in the 1820 's, there were 23 
islands (Bayfield, 1829) . The islands are composed 
of red sandstone in roughly parallel beds, with a 
topsoil of clay and sandy clay. The exposed sand- 
stone is highly susceptible to wave erosion. On 
the north side of some islands and on the mainland 
from the National Park Service Headquarters to Bark 
Point, large sea caves are common. On the southern 
ends of many islands are rocky and sandy beaches, 
sometimes enclosing laqoons. The relief of the 
islands is low and gentle. The highest point, on 
Oak Island, is 480 feet above lake level (Bayfield 
1929; Martin 1965) . 

Except for Madeline Island, the early history of 
the Apostle Islands has not been researched. There 
is ample data on the Indian settlements on Madeline, 
but little on the other islands. Except for an 
Indian sugar grove on Oak Island, tangible evidence 
is lacking that these islands were subject to 
fishing and food gathering activities (GLO notes, 
1852-57). The impact of aboriginal use of the 

islands was sliqht, unlike the effects of the 
Indians' practice of silverculture on the sugar 
maple groves of Isle Royale (Ives 1847). 

Before the land survey period of 1850-57, the 
first settlement by the white man was limited to 
Madeline Island. While this island is not part 
of the National Park System, a brief account of 
man's settlement and activities there is necessary 
to place our story in perspective. Settlement 
began in the late 17th century when a French 
trading post was established on the southwestern 
end of the island. A series of other forts, fur 
trading posts, missions, and dwellings were made 
in the next century and a half, as the French 
interests were succeeded by the British fur 
trading interests. Subsequently the American 
interests predominated. During the American 
period, the American Fur Company established a 
settlement of about 300 people at La Pointe. Be- 
tween 18 37 and 1840, the American Fur Company 
carried on a large fishing enterprise, that em- 
braced all Lake Superior waters from Whitefish 
Pond to Grand Portage (Ross, 1960; Rakestraw, 1968) 

At the time the surveyors began their work 
(1852-57) , Madeline Island had an area of settle- 
ment extending along its western shore, with a 
series of clearings, pastures, farms, dwellings, 
warehouses, and churches. A sizeable amount of 
white and red pine had been cut for construction, 
fuel, and cooperage. The cjtover area had regen- 
erated predominantly to aspen (GLO notes and 
plates, 1952-57; oss, 1960). 

With the exception 
was little trace of ma 
islands during the 185 
Armstrong built a hous 
Oak Island (Armstrong, 
posed about 40 acres ( 
57) . William Wilson b 
some land on Hermit Is 
American Fur Company b 
Stockton Island, near 
plates, 1852-57) . 

of three small areas, there 
n's activities on the 
0's. General Benjamin 
e on the southern tip of 

1927) . The clearing com- 
GLO notes and plates, 1852- 
uilt a house and cleared 
land (Ross, 1960). The 
uilt a small fishhouse on 
Quarry Bay (GLO notes and 

Speculators, promoters, and settlers came with 
the arrival of the railroad about 1870, and the 
extension of the lumbering, mining, and fishing 
frontiers to the area. The cities of Bayfield 
and Ashland were established, and the islands be- 
came the economic hinterland for these booming 
cities (Chappie, 1973; Chappie, 1974; Burnham 
1974) . A large variety of promotional and spec- 
ulative enterprises developed in the area. 

Supported by a National Park Service contract 
to Michigan Technological University. 

Department of Social Sciences, Michigan Tech- 
nological University, Houghton. 

Department of Forestry, Michigan Technological 
University, Houghton 49931. 

A boom period for fisheries 
1870' s. Successor fisheries to 
Company took over its buildings 
cooper ships on Madeline Island 
the railroad reached Chequamego 
fish could be shipped fresh as 
or smoked to Milwaukee, Chicago 
markets. The cities of Duluth 
were established, and provided 
for the fisheries. Fishing bee 

began in the 

the American Fur 
, warehouses, and 
By the 1870's, 
n Bay, and the 
well as salted 
, and Minneapolis 
and of Superior 
venture capital 
ame a major 


occupation at Bayfield 
used, and as time went 
use. The shallow water 
suited to this type of 
were 15 pound net setti 
from Sand River to Bark 
Stockton, and Otter Isl 
settings near their sho 
shipped 310,000 pounds 
Lake trout, 10,000 of s 
and 15,000 of pike. In 
salted 800,000 pounds o 
trout, 31,000 of siskow 
(Smith and Snell, 1891) 

At first gill nets were 
on, pound nets came into 

and sandy bottom were well 
fishing. In 1885, there 
ngs along the coastline 

Point, and Oak, Manitou, 
ands all had pound net 
res. In 1886, Bayfield 
of whitefish, 200,000 of 
turgeon, 40,000 of herring, 

addition, they smoked or 
f whitefish, 305,000 of 
it, and 10,000 of herring 

The end of the frontier period for fishing came 
about 1890, with changes both in organization and 
in technology. The A. Booth Company of Chicago 
moved to Duluth and came to dominate the fishing 
industry on Lake Superior. The "hot head" gaso- 
line engine was invented in Duluth in the 1890 's 
and soon gasoline engine boats began to replace 
the graceful Mackinaw boats. As the Germans and 
Scandanavians moved into the Midwest, herring 
fishing became more important. Norwegian, Finn, 
and Swede-Finns replaced the former French and 
Indian fishermen (Rakestraw, 1968; Chappie, 
1974) . 

Selections from the Ashland Press for 1872 
suggest the development of other commercial 
activities. For example, "Basswood Island is 
doing things in the brownstone and hemlock 
exploitation business. Who says that the re- 
sources and natural advantages of Lake Superior 
are not looking up?" (June 22, 1872). On the 
activities of Mr. Penderqast on Michigan Island: 
"He has a hennery as well as a nursery, and Michi- 
gan Island is prospering." It was noted that he had 
sold 3,000 fruit trees that year (September 28). 
William Knight ran a large advertisement for his 
woodyard on Oak Island: "On the direct channel 
to Duluth, on Oak Island, and a large supply of 
dry hard and hemlock wood is kept on hand" 
(October 12) . 

1900 (Roth, 1898; Lumber Inspector Records, 
1890-1908) . This was the era of animal logging 
and log rafting (Roth, 1898; Wisconsin Lumber 
Inspector Records, 1891-1898). 

After the white pine was logged off, the search 
for Northern hardwoods and hemlock continued and 
tractor logging replaced animal power. One 
operator was John Schroeder, who also brought in 
timber from the north shore and later rafted 
pulpwood. Schroeder conducted railroad logging 
on Outer Island and Michigan Island during the 
1920 's (Wisconsin Lumber Inspector Records, 1891- 
1908; Wisconsin DNR Files, 1959-70; Apostle Is- 
lands, 1970). Since hardwood logs do not float 
well, barges were used to take logs to the main- 
land. After World War II, war surplus landing 
craft were used. Most of the large mills at 
Ashland and Bayfield were closed by this time and 
many of the logs were railed to more distant 
mills . 

Tourism and home building began on the islands 
on a large scale after 1870. Excursions by rail- 
roads and boat lines brought vacationers. Large 
hotels were built in Ashland and Bayfield 
(Burnham, 1974; Chappie, 1974). Excursion boats 
took visitors to the outlying islands, and Made- 
line Island became the home of nabobs from the 
Midwest and East (Ross, 1960). In 1886, Sam 
Fifield established a summer camp, and later 
lodge, on Sand Island. Subsequently, other 
private homes were built on the island. With 
the rise in value of lake property and the 
general prosperity of the period after World War 
II, many of the islands came into private owner- 
ship. After the lights were automated, light- 
house buildings were often leased or bought 
(Chappie, 1973) . 

How have these multitude of activities affected 
the islands and their vegetation over the past 
125 years? The question may be answered first in 
general terms, and then in relation to specific 
groups of islands. 

Quarrying on the islands lasted from about 
1870 to 1900. Four quarries were established for 
obtaining the fine brownstone found on the is- 
lands. Two of these were located on Basswood Is- 
land, a short-lived one owned by John C. Breckin- 
ridge, a Kentuckian who had served as Vice- 
President under Buchanan; and a larger one, owned 
by Milwaukee interests, the larger quarry operated 
from 1869 to 1893. They combined cutting of hem- 
lock for bark with quarrying. On Hermit Island, 
Frederic Prentice in 1891 established a large 
operation with 100 men employed. This quarry 
closed operations about the turn of the century. 
On Stockton Island in 1889 the Michigan Company 
established a large operation which ceased oper- 
ations about 1900. These quarries ceased oper- 
ating because of business and financial problems, 
increased use of steel girders for city con- 
struction, and the growing demand for the light- 
colored Bedford sandstone which lessened the de- 
mand for the dark stone from these island quarries 
(Buckley, 1896). 

The Apostle Islands were exploited for a variety 
of types of timber after 1870. Early cutting in- 
cluded the harvesting of hemlock bark for tanneries, 
the cutting of fuel for steamboats, the use of pine 
and oak for fish barrels, and the clearing of 
forest lands to make room for gardens. About 
1880, the search for white pine began. The is- 
lands were systematically logged for pine and 
most of the merchantable pine had been cut by 

1. Fishing. Every island, including tiny 
Gull Island, has been occupied at one time or 
another by fishermen. For the most part they 
were seasonal visitors, coming in the spring and 
leaving in the winter. There were some exceptions 
to this as both Sand and Rocky Island had year 
round resident fishermen for many years (Chappie, 
1973). Fishermen built cabins, constructed 
docks, wharves, warehouses and storage houses 
for their nets, using local woods. They cut 
wood for fuel, but probably relied a great deal 
on driftwood or logs escaped from log rafts. 
Poles were cut for pound net settings. Most 
fishermen had vegetable or flower gardens, 
fertilized by fish offal. These activities 
were limited in extent, however, and were lo- 
cated on the outer periphery of the islands. 

2. Light 
Island (185 
Sand (1881) 
two to 16 a 
taken up by 
men's estab 
Archives) . 

houses. Five lighthouses were 

on the Apostle Islands: Michigan 
7); Raspberry (1862); Outer (1874); 
; and Devil's (1891). Areas of the 
lighthouse reservations varied from 
cres; a total of about 34 acres was 

these reservations. Like the fisher- 
lishments, these were small, local 
s (Lighthouse Clipping Files, National 

3. Quarrying . The quarrying on Basswood, Her- 
mit, and Stockton Islands had extensive effects 
on the vegetation. Quarrying involves clearing 


vegetation, removing the overburden, and cutting 
out the rock beds, thereby disturbing the 
drainage and altering site stability. It in- 
volved building roads, wharves, warehouses, 
smithys, houses for workers, barns for draft 
animals, and vegetable gardens for the crew. 
These activities had profound effects on fairly 
extensive areas along the edges of Hermit, Bass- 
wood, and Stockton Islands. 

4. Agriculture . Agriculture was common on many 
of the islands, but this practice had a minor 
disturbance effect due to the small acreages 
involved. An exception to this was the 
agricultural activities on Sand Island which were 
extensive and of prime importance. 

5. Lumbering . Lumbering has had the most pro- 
found effect on the islands. The logger searched 
the islands systematically for pine and hemlock, 
then northern hardwoods. By 19 31, Arno Cammerer, 
Associate Director of the National Park Service, 
could report that there were no longer any virgin 
stands of timber on any of the islands except 
Outer Island, which was then in the process of 
being logged off (Cammerer, 1931). Effects of 
this logging depended on several factors: the 
site characteristics, type of stand, type of 
cutting, whether fire followed the operation, and 
whether the land was later used for agriculture. 
Fire generally followed the white pine logging on 
most of the islands. The more xeric sites have 
regenerated back to some pine, but mostly aspen, 
paper birch and tolerant conifers. The mixed 
hardwood-pine and hemlock stands have reverted 

to Northern hardwoods, aspen and paper birch. 
Hardwood logging was a high-grading proposition, 
with the best quality sugar maple, yellow birch 
and other species removed. The result is 
scattered relics of the original stand and poor 
quality second growth. 

Cne additional aspect of the disturbance his- 
tory should be mentioned. With these activities, 
exotic plants were often introduced to the is- 
lands. Apple trees are still found on some old 
agricultural sites. In the era of animal logging, 
grain and hay were brought to the island as fodder. 
Seeds from these germinated, and often flourished. 
Harlan Kelsey, in 19 30, followed a logging road 
on one of the islands and came to a clearing. 
There he found timothy four to five feet high, 
red clover and red top three to four feet high, 
and white clover two feet high (Kelsey, 1930) . 

Some additional examples of the land use his- 
tory may be cited by examining specific islands 
or groups of islands. 

Sand Island may be classified by itself, since 
it is the only island that has had continuous 
occupancy for nearly a century. A Norwegian 
fishermen settled on the island in 1870, and later 
farmed. In 1886 and in 1910, resorts were set 
up — one at the southeastern end of the Island 
and the other at West Bay. Other farmers and 
fishermen came to the Island; roads were built, a 
post office established, and a telephone line run 
to the mainland. A great deal of land was cleared 
for gardens and pasture, and we have records of 
some fires. 

The Island was logged before 1898, again in 
1906, probably removing all the white pine. It 
was logged again in the 1940' s for hardwoods, and 
again in the 1970' s. It is now a complex of 
abandoned fields and forests composed largely of 
nixed torthern hardwoods, aspen and paper birch. 
Some soil erosion has been caused by tractor 
skidding during the most recent operations. 

Stockton and Oak Islands had a similar history 
of exploitation. Oak was a steamboat landing, 
later exploited for pine, and in the 1920' s was 
heavily cut over for hemlock and hardwoods. It 
was burned over in 194 3. Stockton was cut over 
for pine in the 1890' s, had a quarrying operation 
during the same decade, and was cut over for hard- 
woods at a later period. An extensive fire swept 
the south shore in 1929 or 1930. Since 1959, both 
islands have been under protection as part of a 
State Forest, and some recovery from this past 
history has occurred. A cruise of Oak Island in 
the 1960 's indicated only about three million 
board feet of merchantable hardwoods on the Is- 
land. This low volume reflects the past high 
grade type cuttings . 

Hermit and Basswood are both medium-sized is- 
lands with a common history of quarrying activity. 
This activity has shaped the southern half of both 
islands. Both had extensive gardening activity, 
and apple trees are still found on Hermit Island. 
Both were cut over for white pine, hemlock, and 
hardwoods. Field examination of Basswood re- 
vealed a great deal of pulpwood cut, corded, and 
left in the woods. Today there is very little 
high quality timber on either island. 

Michigan and Outer Islands have a common his- 
tory of being logged by railroad in the 1920 's. 
Highgrade logging with conventional equipment 
has followed several times since then. Outer 
also had devastating fires in 1930. The present 
forests consist of scattered sawtimber size culls 
and young sapling and pole size Northern hard- 
woods, aspen and paper birch. Cedar, balsam fir, 
and spruce are common understory species. 

The other islands show a simil 
exploitation and change in veget 
logging. Cat Island is a typica 
Originally called Hemlock Island 
large stands of hemlock, it was 
Frederic Prentice in 1887 for it 
and cedar. It was logged and re 
present, hemlock is no longer an 
cies; the dominant species are y 
hardwoods, with some cedar and b 

ar history of 
ation types after 
1 example . 

because of its 
purchased by 
s hemlock, pine, 
logged. At 

important spe- 
oung Northern 
alsam fir in the 

As Kelsey noted in 19 30, remarkable and far 
reaching changes had occurred. In some 
instances, even greater changes have occurred 
since then. This historical and ecological 
background can be the subject of many specialized 
studies by the historian, and as background for 
future ecological studies. 



Parts I and II. Washington, D. C. 
ARMSTRONG, B. G. 1927. Reminiscenses of life among 

the Chippewa, Part 2. Wisconsin Magazine of 

History, SS. 4 pp. 287. 
BAYFIELD, H. W. 1829. Outlines of the geology of 

Lake Superior. Trans, of the Literary and Hist. 

Soc. at Quebec, Vol. 1, pp. 1-43. 
BUCKLEY, E. R. 1896. On the building and ornament- 
al stones of Wisconsin. Wise. Geol. Surv. Bull. 

10, Madison, Wisconsin. 
BURNHAM, G. M. 1974. The Lake Superior country in 

history and in story. Ashland, Wisconsin 464 pp. 
CAMMERER, A. 1931. Apostle Islands. In: Proposed 

National Monuments, Apostle Islands, RG 48, 

National Archives, Washington, D.C. 
CHAPPLE, J. C. 1973. Apostle Islands source book. 

Ashland, Wisconsin. 8 p. 
CHAPPLE, J. C. 1974. Ashland County, Wisconsin, 

Ashland, Wisconsin. 


Islands, 1852-57. National Archives, RG 48. 

IVES, W. 1847. Survey notes of William Ives, Gener- 
al Land Office Survey. State Historical Collec- 
tions, Ann Arbor, Michigan. 

KELSEY, H. 1931. Report on the Apostle Islands 
National Park project, January 31, 1931. In: 
Proposed National Monuments, Apostle Islands, 
RG 48, National Archives, Washington, D.C. 

LIGHTHOUSE CLIPPING FILE, no date. RG 26, National 
Archives, Washington, D.C. copies at NPS Hdqtr. , 
Apostle Islands National Lakeshore, Bayfield, 

MARTIN, L. 1965. The physical geography of Wiscon- 
sin. Madison, Wisconsin pp. 446-478. 

RAKESTRAW, L. 1968. Commercial fishing of Isle 
Royale. Report to the National Park Service, 
Houghton, Michigan. 

ROSS, H. 1960. LaPointe: Village outpost. Edwards 
Brothers, Ann Arbor, Michigan, 200 pp. 

ROTH, F. 1898. Forestry conditions and interests 
of Wisconsin. Wise. Geol. and Nat. Hist. Survey 
Bull. No. 1. 

SMITH, H. M. , and M. Snell. 1891. A report of the 
fisheries of the Great Lakes in 1885, in: Report 
of the commissioner of fish and fisheries, 1887, 
House Misc. Document 133, 50th Congress, 2nd 
Session, Washington, D.C. pp. 31-70. 

WISCONSIN, DNR FILES, 1959-70. Wise. State Hist. 
Soc. Univ. of Wise. , Madison. 

Wise. State Hist. Soc. Univ. of Wise, Madison. 



Carol A. Jefferson 


The Apostle Islands, covering 7 7 square miles, 
form an archipelago of 22 islands in Lake Superior. 
They consist of glacial till and Pleistocene red 
clays overlying ancient sandstones. Sandy beaches, 
spits, tombolos and cuspate forelands are of more 
recent lacustrine origin. Adjacent lies the Bay- 
field Peninsula, with a 10-15 mile sand covered 
ridge, the Bayfield Hills, which forms the north- 
ern terminus of the Wisconsin pine barrens. The 
dominant vegetation of the wilderness islands is 
second-growth mixed deciduous-conifer forests with 
relict stands of Tsuga canadensis, Betula lutea, 
Pinus strobus and Abies balsamifera - Picea glauca. 
Parallel ridges and runnels, with the runnels 
filled with wetlands, typify the sand regions. 

Previous research on the sand vegetation of the 
Apostles has been minimal. Curtis (1959) listed 
the species of beach and dune communities based on 
two locations in this area, also noting their 
marked difference from dune vegetation along Lake 
Michigan. Murphy (1931) discussed the geography 
of the northwest Wisconsin pine barrens, comment- 
ing specifically on the open nature of the barrens 
and how fire exclusion probably resulted in dense 
growth of Pinus banks iana . Buss (1956) examined 
the plant succession on a sand plain in north- 
western Wisconsin. Concurrent with this present 
study, Barbara Coffin studied the geomorphology 
and vegetational development of the Stockton 
Island tombolo. Results of her study are des- 
cribed elsewhere in these proceedings. An 1872 
map of the LaPointe Light Station on Long Island 
indicates that the vegetation is a forest of 
stunted, scattered Pinus banksiana , interrupted 
by narrow marshes--vegetation similar to that 
occurring on Long Island today. 

Vegetational disturbances include selective 
logging for Pinus strobus and P. resinosa, deer 
browsing, fire and settlement. Today only 
Madeline Island has permanent residents, but both 
Stockton and Long Islands were seasonally occupied 
by Indians who fished and gathered berries (Ross, 
1960). Presumably the Indians burned the woods, 
as they did on the peninsula (Murphy, 1931) , to 
enhance Vaccinium production. Selective logging 
for pines occurred intermittantly between 1870 
and the 1940 's on spits on the islands and penin- 
sula. Nearly every site visited showed evidence 
of logging and subsequent burns. Deer are present 
on several of the islands; their most visible 
impact is the elimination of Taxus canadens is , 
which is preferentially browsed (Beals, 1958). 

The purpose of this study was to identify 
upland and wetland plant communities on sandy 
substrates in the Apostle Islands and to deter- 
mine the successional development and status of 

This study was supported in part by Sigma Xi , 

the National Scientific Research Society. 

Department of Biology, Winona State University, 

Winona, Minnesota. 

the vegetation. An ecological understanding of 
these fragile areas may aid personnel of the 
Apostle Islands National Lakeshore in locating 
and managing recreation areas and campsites. 


During the summers of 1975-1976, 27 stands and 
five line transects were sampled for plant species 
cover. Stands had homogeneous vegetation and were 
generally in excess of four acres. Ground cover 
vegetation was recorded in 28, 20 x 50 cm plots, 
shrubs in one 25/5m plot, and trees in one 25 x 
15m plot per stand. In each stand the largest 
representative of each dominant tree species was 
aged by counting growth rings. Depth of the 
humus in the soil was also recorded. The point- 
quarter method was used to sample along transects. 

An importance value equaling one half of the 
average percent cover and frequency of each 
species in each stand was calculated. Stands were 
then synthesized into communities by the releve 
method. Finally an ordination of communities by 
floristic values and soil development was executed 
End points for the floristic axis represented 
communities on the most xeric and hydric sites. 


brevi ligu 
CI adoni a 
Vacci ni urn 
spica ta ; 
canadens i 
Potamoge t 
Al nus rug 
S p . ; La r i 
and Mgric 

upland and five wet 
from the releve sor 
lata ; Pinus banks i a 
sp.; Pinus resinosa 
sp. - Pter id ium aq 
(Tsuga canadensis ) 
s ; Abies balsamifer 
on sp . (pond) ; June 
osa/Chamaeda phne ca 
x lar ic ini a / Betula 
a gale/Juncus sp. ( 

land communities 

ting: Ammophila 

na/Vaccinium Sp./ 

; Quer cus borealis/ 

u i linum/Danthonia 

- (Betula lutea ) /Taxus 

a; Picea mar i ana ; 

us sp. (marsh) ; 


pumi la / Sphagnum sp.; 

a deflation plain) . 

Table 1 lists the species present in the com- 
munities with importance values in excess of 20. 
These data indicate that the Pinus and Quercus 
dominated communities are closely related floris- 
tically, as are the Alnus and Myrica wetlands. 
The most important factors dictating community 
composition of the wetlands is probably saturated 
soils and Sphagnum . The Pinus and Quercus com- 
munities are both open woods on ridge tops with 
well drained sands and little humus accumulation. 
The Abies community was in all cases 30 years old 
or less and occupied openings in the 
forests and Pinus resinosa stands. With 
rubrum, Abies invades non-reproducing Pinus 
banks iana/Vacci ni um/C ladonia woods, in the absence 
of recent fire and/or heavy growth of I ium. 
The presence of occasional mature Pinus strobus 
trees, many saplings, and frequent stumps, indi- 
cates that this species may be codominant with 
P. banksiana in unlogged forests. Each of the 
forest communities may be invaded by Sphagnum , 
accompanied by elevated water tables, resulting 
in a raised bog moving up the edge of old dunes. 

The ridge and runnel system of sand deposition 
provides a template for studying succession, be- 
cause the ridges parallel to the lake shore and at 
distal ends of spits are the youngest geologically. 



Pinus banks iana 

Potamogeton sp. 

Myrlca gale 



Alnus rugosa 

Pinus resinosa 

Tsuga canadensis- 
Betula lutea 

Abies balsamifera 

Picea mariana 

Larix laricinia 

Site Moisture 

FIGURE 1. Sand plant community succession, Apostle Islands, 
(Communities designated by dominant species.) 

Thus transects running across and along the dune 
system follow historical vegetational development, 
aside from perturbations. The patterns of de- 
velopment in the tombolos are more complex. 
Figure 1 shows succession patterns derived from 
the transect data, combined with the results of 
the f loristic-edaphic ordination. 


Vegetational variations between the peninsula 
and the islands and among the islands can be at- 
tributed to three factors, the presence of deer, 
human activity, and the time of formation of the 
sand landforms, modified by climatic changes and 
the ability of the species to disperse. 

The effects of deer on forest vegetation of 
the Apostle Islands has been extensively discussed 
by Beals (1958) and Beals and Cottom (1960) . The 
deer on Otter, Cat, Stockton, Hermit and Madeline 
islands have virtually eliminated Taxus canadensis . 

Logging and fires have kept Pinus woods on spits 
in the islands and on the peninsula in an early 
successional state. However, recent efforts to 
control fires on Madeline, Sand and Stockton is- 
lands seems to have resulted in the reduction of 
Pinus banksiana and upland Picea mariana regenera- 
tion, and increased colonization of Acer rubrum 
and Abies. Fires have been suppressed on Long, 
Michigan and Raspberry islands by Light Station 
personnel since the late 1800 's. A couple of years 
ago fire burned the dune vegetation on the east 
side of Stockton Island, severely reducing the 
populations of Juniperus communis, Pinus strobus 
and Vaccinium , which are making little or no come- 
back. In the pine barrens region of the penin- 
sula, wild fires have been restrained since the 
establishment of Chequamegon National Forest in 
the 1930' s, with a resultant forestation by dense, 
uniformly aged stands of P. banksiana (Curtis, 
1959) . 

Logging severely decimated stands of Pinus 
strobus and P. resinosa , so that only a few, iso- 
lated stands of mature pines remain. Dating of 
many pines by annual ring counts show that the 
pines are either 100, 60 or 30 years old, thus in- 
dicating the dates of logging. Pinus strobus 
regeneration seems to be the best on the sites with 
less humus accumulation, where p. resinosa is 
regenerating better. 

Quercus boreal i s 
spits of the peninsu 
Islands. All of the 
nent Indian habitati 
been originally plan 
Outer, Michigan and 
solitary oak trees g 
proximity to twentie 
the trees are all le 
individuals may have 
transported by man. 

occurs in abunda 
la and on Madeli 
se areas were si 
on, thus the oak 
ted for acorn pr 
Stockton islands 
rowing on the sa 
th century fishi 
ss than 60 years 
been planted or 

nee only on 
ne and Long 
tes of perma- 
s may have 
oduction. On 

each of the 

nd is in close 

ng camps, and 

old. Such 


On the other hand, Quercus distribution may 
reflect its inability to disperse to sand regions 
on the off shore islands in the time period that 
the sand areas have been habitable for plants. 

Probably the most important determinant of 
forest and wetland communities is the age of for- 
mation and colonization of the sand features. 
Abandoned lake shores, benches and beaches have 
been documented on the Bayfield Peninsula (Cahow, 
1971) and presumably similar features occur in the 
islands. The last continental glaciers retreated 
from this region about 11,500 years ago. During 
the peak of the Wisconsin, Glacial Lake Duluth 
rose to 1085 feet above sea level, covering all 
of the Apostle Islands and much of the Bayfield 
Peninsula, except the Bayfield Hills. Ten thou- 
sand years ago, Glacial Lake Minong occupied the 
present basin of Lake Superior, but water levels 
were at 450 feet, 248 feet below present lake lev- 
els (Farrand, 1969). Since that time the earth 


TABLE 1. Community composition, species importance values >20. 
(Communities indicated by dominant species.) 





















Betula papyrlf era 
Abies balsamif er a 
Pinus resinosa 

Picea mariana 

Amelanchier sp. 
Pinus strobus 
Betula lutea 
Tsuga canadensis 
Acer rubrum 
Acer spicatum 
Pinus banksiana 
Quercus borealis 
Picea glauca 
Juniperus communis 
Taxus canadensis 
Larix laricinia 
Betula pumila 
Alnus rugosa 
Myrica gale 

Chamaedaphne calyculata 
Salix humilis 
Salix interior 
Salix lucida 
Cladonia sp. 
Hudsonia tomentosa 
Commandra richardsonia 
Vaccinium sp. 
Gautheria procumbens 
Gaylussacia baccat a 
Melampyrum lineare 
Danthonia spicata 
Arctostaphylos uva-ursi 
Maianthemum canadense 
Pteridium aquilinum 
Epigea repens 
Clintonia borealis 
Trientalis borealis 
Cornus canadensis 
Deschampsia fTexuosa 
Ammophila breviligulata 
Lathyrus maritimus 
Artemesia caudata 
Sphagnum sp. 
Juncus ef fusis 
Hieracium aurantiacum 
Carex rostrata 
Calla palustris 
Carex lasiocarpa 
Vaccinium oxycoccu s 
Juncus gerardi 
Scirpus validus 
Juncus canadensis 
Equisetum arvense 
Oenanthera biennis 
Potamogeton illinoensis 
Potamogeton natans 
Nuphar variegat n 
Brasenla schreberi 
Utricularia intermedia 

-t- + 


+ + 

+ + + 

+ + + 

+ + 

+ + + 

+ + + 

+ + 

+ + + 

+ + 

+ + + 

+ + + 

+ + + 




has slowly rebounded, and new shoreline sand 
deposits undoubtedly developed with the falling, 
and later, rising lake levels. 

A study of the present lake depths among the 
islands indicates that when the lake was 150 feet 
shallower, the islands were interconnected in 
three clusters, with Outer, Michigan and Stockton 
islands isolated by themselves. The clusters con- 
sisted of Chequamegon Point - Long Island - 
Madeline Island; Basswood - Hermit - Oak - Otter - 
Rocky - Devil's - North Twin - South Twin - Cat - 
Bear - Raspberry - Sand - York - Eagle islands and 
the peninsula; and Manitou Island - Ironwood Island. 
This depth was chosen because inter-island similari- 
ties in sand vegetation closely follow these clus- 
ters and because the lake may have been at this 
level by the time the climate had warmed enough to 
support vegetation similar to today. Even during 
the lake's lowest post-Pleistocene level, Michigan 
Island probably was not connected to the rest of 
the islands . 

Despite the limitations of 
these postulated island cluste 
may account for the presence o 
umbellata on Long Island, the 
Island; Quercus on Long Island 
and the peninsula, and the abs 
banksiana , Picea mariana, and 
Island. The wetland vegetatio 
more homogeneous than the fore 
diversity of strand line plant 
distance from the peninsula 

this analytic method, 
rs and isolations 
f Chima philia 
peninsula and Outer 

Madeline Island 
ence of Pinus 
Larix on Michigan 
n on the islands is 
sts, except that the 
s decreases with the 

Older deposits seem to be 
coarse grained sand, with a 
forested by Tsuga-Betula and 
while the adjacent depressio 
deposits occur on the inner 
lands and tombolos, and were 
exposed sand flats that blew 
as the lake level dropped, 
stabilized by plants as the 
haps 4000-5000 years ago. 

indicated by a red, 
deep humus layer, 

Abies communities, 
ns are bogs. These 
edge of cuspate fore- 

probably formed from 

inland, forming dunes, 
The dunes were later 
climate warmed, per- 

To gain a precise picture of the rates of 
deposition and colonization of the sand, dating of 
bog sediments and determination of historical 
lake levels will be necessary. 

A comparison of forest and wetland plant commu- 
nities of the Apostle Islands to similar communi- 
ties in northern Wisconsin, as described by Curtis 
(1959) and others, indicates that the island 
communities have lower species diversity, especially 
in hepatics and deciduous trees. The island re- 
gion is at the northwest limit of the continuous 
Tsuga forest and several of its species. Curtis 
(1959) noted that, as in the islands, Abies in- 
vades openings in Tsuga and Pinus strobus forests 
occupying rich soils. Wetland communities of the 
islands are lower in species diversity. 

The most striking difference between the sand 
vegetation of the Apostle Islands and northern 
Wisconsin is the lack of Mgrica aspleni folia as a 
dominant shrub in the pine barrens. This species 
does not occur on the islands, nor occurs on any 
peninsula spits, although it is abundant only 
5-10 miles inland. The explanation for this 
distribution may be lack of a micronutrient or 
symbiont. M. aspleni fol ia is a nodulated nitro- 
gen fixer, as is m. gale which is abundant on de- 
flation plains and the edges of bogs in the 
islands. Levy's (1970) list of plant species in 
the openings in the sand barrens has 74% of the 
prevalent species in common with the Pinus, 
Quercus and Ammophila communities in the islands. 


This study indicates that the sand vegetation 
of the Apostle Islands is similar to vegetation 
elsewhere in northern Wisconsin, but has fewer 
species. Discontinuities in species distribution 
between the islands may be the result of histori- 
cal connections and barriers. Seven upland and 
five wetland communities were identified, with 
upland vegetation culminating in Tsuga canadensis- 
Betula lutea or Abies balsamifera forests, and 
wetlands becoming Sphagnum bogs. 


BEALS, E. 1958. The phytosociology of the Apos- 
tle Islands and the influence of deer on the 
vegetation. M.S. Thesis, University of Wis- 
consin, Madison. 

BEALS, E. and GRANT COTTOM. 1960. The forest 
vegetation of the Apostle Islands, Wisconsin. 
Ecology 41:743-751. 

BUSS, I. O. 1956. Plant succession on a sand 
plain, northwest Wisconsin. Trans. Wis. Acad. 
Sci. Arts Lett. 45:11-20. 

CAHOW, A. C. 1971. Abandoned shorelines and re- 
lated geomorphic features of Bayfield County, 
Wisconsin. M.S. Thesis, Michigan State Univer- 

CURTIS, JOHN T. 1959. The vegetation of Wiscon- 
sin: an ordination of plant communities. Uni- 
versity of Wisconsin Press, Madison. 657 p. 

FARRAND, WILLIAM R. 1969. The Quaternary his- 
tory of Lake Superior. Proceedings of the 12th 
Annual Conference on Great Lakes Research, 
pp. 181-197. 

LEVY, GERALD. 1970. The phytosociology of north- 
ern Wisconsin upland openings. Amer. Midland 
Natar. 83:213-237. 

MURPHY, R. E. 1931. Geography of the northwest 
pine barrens of Wisconsin. Trans. Wis. Acad. 
Sco. Arts Lett. 45:11-20. 

ROSS, HAMILTON H. 1960. LaPointe--village out- 
post. University of Michigan Press, Ann Arbor. 
200 p. 



Barbara Ann Coffin 


The origin of the ridges and troughs that form 
a double tombolo, two bars of sand joining Presque 
Isle Point to the main body of Stockton Island, in 
the Apostle Islands region of western Lake 
Superior is related to the history of water level, 
currents, and wind. The inception of the Presque 
Isle tombolo began approximately 5500 years ago, 
as the lake level began to retreat from the Lake 
Nipissing stage, 185.3 m (608 ft) above sea level. 
A series of parallel ridges were formed as the 
water level lowered, culminating with a final 
barrier beach formed on the northeast shoreline 
of the tombolo. This outer beach is identified 
as an Algoma beach formed between 3200 and 2500 
B.P. Sediment and radiocarbon analyses support 
this interpretation. 

The diverse topography of the tombolo provides 
habitats for distinct vegetation communities. 
Four different topographic zones were identified: 
active dunes, interdune hollows, stabilized 
dune/beach ridges, and inter-ridge troughs or 
filled lake basins. Single vegetation plots were 
studied in the sand dune and bog communities of 
Presque Isle tombolo accordinq to the Braun- 
Blanquet releve system. Releves were arranged 
according to species occurrence. Organized in 
this manner, the releves fall into three major 
groups: hydric, mesic, and xeric. Boundaries 
between adjacent plant communities are visibly 
distinct, but when tabulated they represent a 
nearly continuous gradient from hydric to xeric 
communities. By measure of structure and compo- 
sition, five different vegetation types are 
recognized from this gradient: lake dunes, 
northern conifer forest, northern mixed forest, 
shrub carr, and open bog. 

A combination of topographic zone, vegetation 
type, moisture status, degree of exposure, and 
dominant species define twelve major plant 
communities. Topography and moisture status are 
key factors influencing the distribution and 
composition of plant communities. On a broad 
scale, the age of substrate is not so important 
as water level in understanding successional 
relationships on the tombolo. Within each 
community, cyclic patterns rather than linear 
patterns characterize successional development. 

Presque Isle tombolo demonstrates the dynamic 
relationship between physical and biotic processes 
in natural ecosystems. 


Parallel dune and beach ridges separated by 
linear bogs form a double tombolo, consisting of 
two bars of sand and gravel joining together what 
were formerly two islands. The Presque Isle 
tombolo stretches approximately 2 km between the 
main body of Stockton Island and the point of 
Presque Isle (Figs. 1 & 2) . Stockton Island is 

part of the Apostle Islands archipelago, located 
off the north-central shore of Wisconsin in the 
western half of the Lake Superior basin. This 
area of diverse topography displays a hetero- 
geneous array of distinct plant communities, in- 
habiting a landform created by sediments deposit- 
ed and rearranged by the interaction of wind, 
currents, water level fluctuations, and lake ice. 
Physical and biotic factors create gradients 
along which species align in a continuum from 
abundant to absent. This paper demonstrates how 
geomorphic processes and geologic events continu- 
ing since glacial times have interacted dynamic- 
ally with vegetation development. 


Stockton Island is now administered by the 
National Park Service as part of the Apostle 
Islands National Lakeshore, established in 1970. 
Previous to its inclusion within the National 
Lakeshore, it was administered by the Wisconsin 
Department of Natural Resources; it was recognized 
as a unique area of scientific value (Tans & Read 
1975) . 

The Apostle Island National Lakeshore is 
classified as a recreation area of the National 
Park Service. Plans for the development and 
management of Presque Isle Point at this time 
includes construction of campsites (summer 1976) 
and maintenance of a trail (already in existence) 
on the southwest ridge of the tombolo, bordering 
the lakeshore. On the northwest side of the 
tombolo there is to be no camping, only foot 
travel. The wetlands of the area, lagoons and 
floating bog mats, are so difficult to traverse 
on foot in summer that their deterioration by 
visitors is not anticipated. The area is largely 
inaccessible from the mainland in winter because 
of shifting ice packs. 

It is hoped that further development will not 
occur. The sandy substrate of the ridges is an 
easily erodible surface, part of a delicate 
vegetation-environment system, which under heavy 
human traffic could be drastically altered. Be- 
cause it is an unusually well developed example 
of a double tombolo and supports a unique array 
of dune, bog, and forest communities, it is of 
considerable importance in the preservation of 
ecological diversity and as a study area for 
investigators and observers of natural history. 

Department of Botany, University of Minnesota, 
St. Paul 55108. 

To faci 
tombolo fo 
upon which 
a knowledg 
necessary . 
directly r 
Islands an 


litate an accurate understa 
rmation as the physiographi 

the existing vegetation ha 
e of the physical environme 

Specifically relevant to 
ing is a reconstruction of 
, and climatic events of th 

attention is given to the 
elated to the formation of 
d subsequently Presque Isle 

nding of the 
c foundation 
s developed, 
nt is 
such an 
the glacial 
e Quaternary, 
the Apostle 


Geologic history of the Lake Superior basin. — 
The Lake Superior basin is composed mostly of 
Precambrian rocks, the Jacobsville Sandstone. 
The basic structure of the basin was formed in 
Late Precambrian time when folding and faulting 
created the synclinal structure that exists today. 
The bedrock exposed in the Apostle Islands is the 
red quartzose unfossiliferous Bayfield Sandstone 
(Thwaites 1912). Within the islands, bedrock 
exposures are generally restricted to the northern 
and eastern shorelines, where cliffs carved by 
wave action protrude from beneath a mantle of till 
and lacustrine clays (Engstrom 1972) . 

Glaciation during the Pleistocene was a major 
factor shaping the present-day topography in the 
Lake Superior basin and adjacent areas. The most 
recent advance of the continental ice sheet, about 
12,000 years ago, covered the entire basin of Lake 
Superior. As this ice sheet retreated northward 
a series of lakes was formed from the meltwater. 
The shorelines cut by these meltwater lakes 
provide evidence for the reconstruction of the 
Superior basin lake-level history. Interpreting 
this evidence is complicated by the following 
considerations: 1) the effect and extent of post- 
glacial crustal uplift on the present elevations 
of past shorelines, 2) positions of 'hinge lines", 
north of which rebound has occurred, 3) submergence 
by flooding of shorelines south of the "hinge lines' 
and 4) the elevations of potential outlets 
utilized by the meltwater lakes during their 
existence. As evidence on each of these is 
accumulated through new research, previous 
hypothetical explanations must be altered to con- 

Because of uplift of the North Bay outlet, flow 
was transfered to the Illinois River outlet and 
the Port Huron outlet (still in use today between 
Lakes Huron and Erie) , and the lake level again 
began to fall as these outlets eroded. The lake 
level fell approximately three meters to the Lake 
Algoma stage, 182 m (596 ft) MSL, where it 
stabilized during a halt in the erosion of the 
Port Huron outlet. Renewed lowering formed the 
minor Sault stage, with shorelines identified at 
178.6 m (586 ft) MSL (Farrand 1960). This trend 
ceased about 2200 years ago, when glacial rebound 
caused uplift of the bedrock sill at Sault St. 
Marie, raising the water level in the Superior 
basins (Farrand 1969) . Rebound has continued, 
accounting for the gradual rise to the present 
level of Lake Superior at 193.6 m (602 ft) above 
sea level (USDC 1976) . Since Lake Superior 
emerged as a separate lake, rebound has progres- 
sively elevated the northeast shore, and flooded 
the shores that lie south of the isobase of zero 

The general topography o 
tells much of how they fit 
tory of the Lake Superior b 
ice sheet moved across the 
direction, parallel to the 
the islands and the channel 
Presumably, old glacial val 
and subsequent isolation of 
in combination with erosion 
(Collie 1901). Thus a chan 
important implications for 
destruction of islands and 
islands . 

f the Apostle Islands 
into the glacial his- 
asin. The most recent 
islands in a NE-SW 
long axes of many of 
s between them, 
leys were flooded, 
high land surfaces, 
, formed the islands 
ge in lake level has 
the construction or 
connections between 

Study of the glacial history in Lake Superior 
began as early as 1850 with the expedition of 
Louis Agassiz. The descriptions by Agassiz were 
followed by more detailed study of glacial features 
by Taylor (1894), Leverett (1929), Sharp (1953), and 
Hough (1958). Farrand' s research (I960) on former 
shorelines has presented the most recent acceptable 
hypothesis for understanding Quaternary history 
(Fig. 3). Additional research and radiocarbon 
dates have since somewhat altered Farrand 's 
original time scale (Lewis 1969, Saarnisto 1975). 
However, Farrand must be credited with building a 
framework that can be modified and strengthened as 
future research dictates . 

The western Lake Superior basin was filled with 
glacial ice for the last time during the Nickerson 
phase of the Superior Lobe, approximately 12,000 
years ago (Wright and Watts 1969) . The glacial 
meltwater lakes first formed following retreat of 
this ice cover, flowed south through the Moose 
River and then through the Brule River outlet to 
the Mississippi River (Fig. 4). With further re- 
treat, lower outlets in the upper peninsula of 
Michigan were opened, and the water flowed east 
into the Michigan and Huron basins (Farrand 1960) . 
During this period much of the lacustrine sediments 
that now cover the islands was probably laid down. 
The lake level became progressively lower as out- 
let channels were freed of ice. This trend ended 
with the low Houghton stage (8000 B .P) when the 
water level stood only 100 m (360 ft) above mean 
sea level (MSL) . At this time waters from Huron, 
Michigan, and Superior basins drained northeast 
through the North Bay outlet to the Cttawa 
River (Fig. 4) . Glacial rebound caused uplift 
of this principal outlet, and lake levels began 
to rise, culminating at 185.3 m (608 ft) MSL, 
the Lake Nipissing stage. It is now believed 
that the Nipissing high level occurred 
approximately 5500 years ago, contrary to the 
traditional date of 4000 years B.P. (Lewis 1969, 
Saarnisto 1975) . 

Geomorphology of tomb 
is Italian in origin, de 
examples of these featur 
coast (King 1959) . Alth 
studied feature, tombolo 
Italy, the classic Orbet 
1899) , New England (John 
Islands, Quebec, and Nor 
(Farquhar 1967) . Tombol 
bination of factors : Ion 
drifting, winds (here in 
waves, and water-level c 
originally formed by the 
spit or by the convergen 

olos. -- The term tombolo 

scribing the classic 

es found on the Italian 

ough not a commonly 

s have been described in 

ello tombolo (Gulliver 

son 1919) , and Magdalen 

th Cape, New Zealand 

os are created by a com- 

gshore currents , beach 

combination with ice), 
hange . They are 

development of a single 
ce of two or more spits. 

The direction of prevailing and storm winds 
are basic to understanding the origin of a tombolo. 
The waves (and here ice) are the agents that 
transmit wind energy into the kinetic energy that 
builds spits and tombolos. The waves most 
effective in building underwater bars are those 
that come from the direction of maximum fetch 
(Farquhar 1967). Longshore currents, created by 
waves meeting shorelines obliquely, cause 
deposition of underwater bars of sand (King 
1959) . Sediments carried by these currents are 
deposited at right angles to the direction of 
dominant on-shore wave action. 

The way in which underwater bars emerge to 
form spits or tombolos is controversial. 
Johnson (1919) states that wave action alone is 
capable of building a submerged bar above a 
water surface of constant elevation. .Accord- 
ing to Evans (1972), however, subaqueous ridges 
are not built above a water surface of constant 
elevation by the work of waves depositing 
material from the bottom. Lewis (1931, 1932) 
states that the emergence of spits is the result 
of deposition by obliquely approaching waves 
during major storms. The importance of ice 
action on lake shore is examined by Scott (1926) . 


He explains the ramparts found on the shores of 
lakes to be formed in part by "ice- jam", the 
movement of beach material by ice shoved on shore 
by wind and wave action. These hypotheses all 
become of secondary importance if there are major 
fluctuations in water level. A falling water 
level will expose previously submerged ridges. 

In the case of the Presque Isle tombolo, two 
separate spits extended from the main body of 
Stockton Island enclosing a triangular lake (now 
bog) . Convergence of these spits (S a , S^ of 
Fig. 2) joined Presque Isle Point to the main 
body of Stockton Island, forming what is described 
as a double tombolo. Farquhar (1967) states that 
double tombolos are formed by waves that approach 
the tied islands at different angles on the lee 
side. This is in agreement with Engstrom's 
(1972) findings that cuspate forelands, spits, 
and tombolos have a preferred orientation in the 
Apostle Islands, the southern (lee) side of the 
islands. The Presque Isle tombolo projects 
southward from the lee shore of Stockton Island. 
The action of prevailing winds (see Climatology) 
and particularly storm winds from the northwest 
and northeast in the fall and from the southwest 
in the summer can account for the development of 
two separate spits at this site. 

Climatology . -- The climate of the Apostle 
Islands is basically continental, moderated by 
the presence of Lake Superior. At Ashland, a 
coastal town 48 km (30 mi) south of Presque Isle 
Point. Winters are cold with average January 
temperatures of -10.6°C (13°F) , and summers are 
cool with average July temperatures of 19.4°C 
(67°F) . Average annual precipitation is 70 cm 
(29 in) (USDC 1974a) . 

Climate parameters of particular importance to 
the study of processes involved in tombolo forma- 
tion, include prevailing wind directions and 
duration and extent of ice cover. The closest 
source of compiled wind data is Duluth, 
Minnesota, approximately 100 km (63 mi) west of 
Presque Isle (USDC 1974b) . Winds of Beaufort 
force 3 and greater, capable of producing waves 
of significant dimensions, were extracted from 
the Duluth statistics. Prevailing winds during 
the months of Hay, June, and August are easterly. 
Beginning in September and continuing through 
April, and also in July, prevailing winds are 
from the west and northwest. 

Storm winds 
southwest, and 
The most severe 
In this season 
tropical air ma 
energy from the 
create cyclonic 
"northeasters" . 
with winds from 
kph) , later shi 
45 to 60 mph) , 
coming straight 
Minnesota north 

in the summer are from the west and 
during the autumn from the northeast, 

weather occurs during November, 
the intensely different polar and 
sses converge and, assisted by heat 

relatively warm unfrozen lake, 

storms, commonly referred to as 
Typically these storms begin 

the SE (35 to 40 mph, or 56 to 64 
fting to the NW (72 to 96 kph or 
and finally veering into the NE 

down the lake parallel to its 

shore (Loy 1962) . 

The extent of ice cover is important because of 
the potential for movement of beach sediments by 
ice packs. Estimates of ice-cover duration are 
made from the USDC (1976) and Marshall (1967). 
Navigation for Ashland harbor is closed on the 
average from December to April. Marshall's 
statistics suggest that the duration of ice cover 
is greater in the channels separating inner is- 
lands than those separating outer islands. The 
northeast side of the Presque Isle tombolo, due 
to its location in the eastern outer limits of the 

Apostle Islands, is exposed during most of the 
winter to the action of wind-driven ice packs. 


The Apostle Islands and northern Wisconsin are 
classified as part of the northern-hardwoods 
forest, described by Kuchler (1964) as tall, 
broadleaf deciduous forests with an admixture of 
needleleaf trees. Eeal and Cottam (1960), who 
studied the forests of the Apostle Islands, agree 
that they contain representative forests of 
northern Wisconsin but recognize some important 
differences . Most salient of these is the 
scarcity of well-drained habitats . Upland areas 
on the islands usually resemble lowland areas on 
the mainland. Beal and Cottam (1960) found the 
dominant trees in the upland forests to be 
Betula lutea (Yellow Birch), Thuja occidentalis 
(Northern White Cedar), Betula papyri fera (Paper 
Birch) , and Acer saccharum (Sugar Maple) . Sand 
spits when sufficiently elevated above the water 
table are dominated by Pinus resinosa (Red Pine) 
and Pinus strobus (White Pine). The bog 
communities are typical of northern Wisconsin, 
dominated by ericaceous shrubs, sedges, and 
sphagnum moss, and support in some areas a 
scattered population of Picea mariana (Black 
Spruce) and Larix laricina (Tamarack). 

A survey of the vegetation of Stockton Island 
(Annonymous 1974) identifies Yellow Birch-Maple 
and pure Maple stands as the forest type dominant 
on the main body of the island. Presaue Isle 
Point, once a separate island, is described as a 
mesic area of mixed conifer and hardwood species. 

The Presque Isle tombolo supports a composite 
of sand dune and bog communities. The sand-dune 
communities are found in varying states: actively 
aggrading dunes and stabilized dunes (supporting 
either a xerophyllous population of low herbaceous 
and shrubby plants or covered with a forest 
canopy) . The big communities include a filled 
lake basin with firm sedge-sphagnum mats support- 
ing trees, narrow strips of wet sphagnum bog fil- 
ling troughs between old beach ridges, and open 
ponds with long, linear sedge-sphagnum mats. 

Beach and sand dune communities have been the 
object of considerable study world wide. Dune 
environments provide ideal laboratories for the 
study of dynamic change in a natural system. 
Cowles (1899) in his classic work in the Lake 
Michigan dunes looked at the relationship between 
vegetation and physiography in terms of succession- 
al development. He found that a dune system 
functioned as an excellent habitat for observation 
of plant formations that are rapidly passing into 
other types by reason of changing environment. 
The interrelation of physical and biotic factors 
was continued, in the Lake Michigan dunes, in more 
specialized studies by Olson (1958 a, b, c, & d) 
and Lang (1958) . Studies of similar focus have 
been conducted on both the Atlantic and Pacific 
coasts. Some of the more comprehensive studies 
include Oosting and Billings (1942), Cooper 
(1958, 1967), Martin (1959), and Au (1974). 

Bogs have been studied extensively in glaciated 
landscapes in many locations over the entire 
northern hemisphere. Bogs are of remarkably 
uniform structure and composition throughout 
circumboreal regions. The term bog, as defined 
by Curtis (1959) , refers to a soil-vegetation 
complex composed of a specialized group of herbs 
and low shrubs growing on a wet, acid soil of 
peat. The two f lowering-plant families most 
commonly encountered are the Ericaceae and 
Cyperaceae, which create a canopy over a ground 
covering of sphagnum moss. Studies of bogs in 
the adjacent states of Minnesota and Michigan 


include those of Conway (1940) , Gates (1942) , and 
Heinselman (1970) . 

The dual 
Isle tombol 
both vegeta 
biotic and 
was conduct 
1975 and 19 
summer expl 
cal survey, 
coring and 
were comple 


approach to unde 
o requires method 
tional and topogr 
physiographic pro 
ed during parts o 
76. The studies 
ored the present 
by reconnaissance 
During the seco 
systematic analys 

s which el 
aphic patt 
cesses . F 
f both the 
and gener 
nd summer 
is of the 

the Presque 
ems , and 
ield work 

summers of 
the first 

al ecologi- 

Most important in the interpretation of 
topographic features was a careful study of the 
literature on water-level history of Lake Superior. 
To provide verification and further evidence for 
this interpretation, transects were established in 
the large triangular bog (B of Fig. 2), along 
which probes were made and cores were taken 
systematically. A 4. 8-cm-diameter Livingstone 
piston sampler (Cushing and Wright 1965) was used 
for coring when a continuous sample was desired. 
A 1. 6-cm-diameter Davis sampler (Wright et al. 
1964) was used for collecting data on depths and 
contact zones of sediments in the bog basin. 
Content of organic matter was determined by loss 
on ignition for representative samples from eight 
meters of sediment. A composite of five peat 
samples from 6.54 m depth at station 18 on tran- 
sect A was dated by radiocarbon analysis by 
Geochron Laboratories, Cambridge, Mass. 

A topographic and ve 
(Fig. 7 & located on Fi 
across the long axis of 
at right angles, 30 °N o 
measured from the lake- 
1975 by use of an Abne 
water table under the h 
measured (August 31, 19 
with a flattened tip an 
(to admit ground water) 
described by Turnoch an 
the tube had been in pi 
boo stick attached to a 
was lowered until it to 
and the depth below the 

getation profile 140 m long, 
g. 8 B-B') was laid out 

the tombolo approximately 
f east. Elevations were 
level datum on August 30, 
hand level. Depth to the 
ighest dune ridge was 
76) by driving down a tube 
d perforated lower portion 
, according to the technique 
d Lawrence (1953) . After 
ace 24 hours, a split barn- 
calibrated nylon thread, 
uched the water surface, 
soil surface could be 

The vegetation analysis, described not only 
composition but the structure of the community in 
relation to the physiography of the habitat. 
The following techniques for gathering information 
were utilized: (1) aerial photography (Abrams 
Aerial Survey, Lansing, Mich. July 1974) and a 
personal flight (July 1976) , (2) reconnaissance on 
foot, (3) compilation of releves according to the 
Zurich-Montpellier school of phytosociology 
(Braun-Blanquet 1932) at twenty-five locations, 
(4) symbolic representation of vegetation on a 
topographic profile (Fig. 7), (5) construction 
of a vegetation map (Fig. 8), and (6) collection 
of voucher specimens (identified and deposited 
in the University of Minnesota Herbarium) . 
Plant nomenclature follows Gleason and Cronquist 

Because of the high degree of homogenity with- 
in a given unit, boundaries of plant communities 
are in most cases clearly demarcated in relation 
to topography. Releves were located in each 
major topographic unit on the tombolo. It was 
ascertained by testing with sample plots of 
several sizes that a 10 x 40 m area was sufficient 

for sampling the vegetation of each unit. This 
size was modified in some areas, for example some 
of the beach ridges were not 10 meters in width. 
For each releve, all species of vascular plants 
and abundant non-vascular genera were recorded on 
the Braun-Blanquet cover-abundance scale (Table 
1) (Mueller-Dumbois and Ellenberg 1974) . 
Additionally, details of slope, aspect, altitude, 
height of vegetation, soil characteristics, and 
physiographic features were noted in a full 
description of the releve site. 


Origin of Presque Isle tombolo 

The model of lake- level history in the Lake 
Superior basin worked out by Farrand (1960, 1969) , 
modified by Saarnisto (1975) , and here summarized 
in the introduction, is used to interpret the 
origin and development of the Presque Isle double 
tombolo. It is believed by the present writer 
that this geomorphic feature developed as the 
lake level began to retreat from the high Nipissing 
stage, 188 m (608 ft) MSL. The high period for 
this stage was approximately 5,500 years ago 
(Saarnisto 1975) . Thus the two forested dune 
ridges (S a , Sv, of Figs. 2 & 5) , which enclose the 
large triangular-shaped bog (B of Figs. 2 & 5) , 
were formed as the lake retreated from the 
Nipissing high level. 

It is likely that the dune ridges were super- 
imposed on exposed underwater ridges of sand, 
which had been formed as the lake level rose 
from the low Houghton stage. The rising lake 
level may have caused increased erosion of adja- 
cent bedrock shorelines, augmenting the amount of 
material available for deposition of underwater 
bars. A falling water level is one explanation 
for the emergence of underwater bars between 
Presque Isle Point and Stockton Island. The other 
alternative is that underwater bars emerged prior 
to the water-level drop because of other factors: 
high storm waves, ice push, or gradual deposition. 
In the latter alternative, a falling water level, 
which would have followed the establishment of 
the emerged bars (spits) , would merely have ex- 
posed new material and initiated further growth 
of the existing spits. Thus the emergence of 
underwater bars prior to or concurrently with the 
falling water level does not affect the subsequent 
development of the emerged surface. Presque Isle 
tombolo probably formed as the falling water level, 
aided in beach drifting and longshore currents, 
caused elongation of the emerged spits, eventually 
connecting Presque Isle Point to Stockton Island. 
Dunes would have developed on these spits as in- 
creasing amounts of beach material were exposed 
to wind action by lower lake levels . 

As the water 
ridges (S c of F 
northeast side, 
that mark the f 
during major st 
of these proces 
The absence of 
of the tombolo 
magnitude and f 
strike that sho 
of the waves is 
shore than the 
are evidence th 
on the northeas 
structures, whe 
on the southwes 

level continued to fall, beach 
igs . 2 & 5) , still visible on the 

were formed. The beach ridges 
ailing lake level were deposited 
orms by ice push or by combination 
ses and the retreating water level, 
beach ridges on the southwest side 
is attributed to the lesser 
requency of storms and winds that 
reline. Additionally, the fetch 

considerably less on the southwest 
northeast shore. The beach ridges 
at the wind and wave conditions 
t shoreline can build shore 
reas the wind and wave conditions 
t shoreline do not possess this 


FIGURE 1. Map of Presque Isle tombolo and its loca- 
tion in the western half of the Lake Superior basin. 




a O C O O 






2 — 

- ■" 




^ •" 




^. -- 


X oo 






Z <T> 














=0 *. 









FIGURE 2. Vertical aerial-photograph of Presque Isle 
tolbolo S a & Sj-, - forested dune-ridges, S c - paral- 
lel ridges and troughs S<j - barrier beach/dune ridge, 
and B - large triangular-shaped bog. 

FIGURE 3. Profile illustrating approximate 
elevation, age, and outlet of the post-glacial 
lakes of the Lake Superior basin (adapted from 
Farrand 1960) . 

FIGURE 4. Map showing the outlets used by melt- 
water lakes formed by the retreating Wisconsin 


dune deposition S, t S D 

FIGURE 5. A reconstruction of the historical 
development of the Presque Isle tombolo. 


Stratum Symbol 

T - tree layer (height greater than 5 m) 

saplings (height I to 5 m) 

seedlings (height less than I m) 
S - shrub layer 

S. - (woody plants less than 1 m) 

S. - (woody plants I to 5) 
H - herbaceous layer 


gram I no Ids 
G - ground layer 


I ichens 

Rating Symbol 
cover degree - 


r - single occurence 

t - occasional, cover less than 5$ 

1 - plentiful, cover less than 5$ 

2 - very numerous, cover 5-25$ 

3 - any number of Individuals, cover 25-50$ 

4 - any number of Individuals, cover 50-75$ 

5 - any number of individuals, cover 75-100$ 

40 60 

percent orgamcs 

" 100 

FIGURE 6. Percent of the organic matter in bog 
sediments as a function of depth from surface. 

TABLE 2A. Summary chart of species present in the tree layer 
and shrub layer in 25 releves on Presque Isle tombolo. 

Releve number 

total no. of species 



poorly drained 

4 3 j 22 2 

2S 20 1? 2S1321 11 e 9 24 18 IB 14 12 10 8 23 5 19 
'4 2118 8 2016 IS IS 1224 8 14 14 IS 19 IS 23 30 22 2110 2124 8 9 


Pinue strobue 
Betula papyrifera 
Piaea mariana 
Pinue reeinoea 
Acer rubrum 
Abies baleamea 
Sorbue americana 
Tonga canadensis 
Pinue banksiana 
Thuja occidentalie 
Betula lutea 
Populus grandidentata 
Larix laricina 

Shrub II 
Amelanchier ep. 
Alnus rugoea 
Phyeocarpue opulifoliue 

Shrub I 

Chamae daphne calyculata 
Myrica §ale 

Andromeda glaucophylla 
Kalmia polifolia 
Vaccinium macrocarpon 
Vaccinium Oxycoccue 
Ledum groenlandicum 
Vaccinium angustifolium 
Gaylussaccia baccata 
Vaccinium myrtilloides 
Gaultheria procumbene 
Gaultheria hiepidula 
Juniperue communis 
Spirea alba 
Salix humilie 
Prunue pumila 
Aratoetaphylos uya-urei 
unknown if 4 9 

Pinue Strobue 
Betula papyrifera 
Picea mariana 
Pinue reeinosa 
Acer rubrum 
Abies baleamea 
Thuja occidentalie 
Larix laricina 

1 r 2 2 2 

+ 2 

1 1 + + 12 2 
2 4 5 + 

1 + 2 3 

+ r 1 2 1 

+ + 1 

r + 

+ 3 + + + 

1 + r + 

1 + + 

13+ + 


+ 3 

2 * 










+ 3 

2 + 











r r 



















1 1 


♦ 1 




2 2 















+ 2 

+ 2 + 




1 + 




+ 11 






1 + + + 

+ + + + 


+ + 



Sedae Hoi low 


Black Spruce 




Ericaceous ShruPs- 

FIGURE 7. Profile of Presque Isle tombolo showing the relationship between topographic zones 

and plant communities. 


TABLE 2B. Summary chart of species present in the herb layer 
(forbs) in 25 releves on Presque Isle tombolo. 




Releve natter 25 20 17 15 1321 11 
total no. of species 24 21 18 8 20iei5 

6 9 24 
1513 24 

18 16 1412 10 8 23 5 19 
8 14 1415 19 16 23 3022 


4 3 1 22 2 
1011 24 8 9 


Herbs Iforb.) 
Sarracenia purpurea 
Droeera rotundifolia 
Calopogon pulchellue 
Menyanthea trifoliata 
Pogonia ophiogloeeoidee 
Iris versicolor 

II + + 2 + 
ill 11 
11+++ + 

III 12 

111 1 + 
+ + 


r 1 + 


Drosera anglica 
Smilacina trifolia 
Potentilla palustrie 

111 1 

+ + 



Unknown U8 
Trientalia borealie 



( + 





Lysmmachia terrestrie 
Eriocaulon eeptangulare 
Oenothera biennis 



Potentilla norregica 
O&munda cinnanomea 
Pteridium aquilinum 


3 + 

♦ 1 1 





Haixvxtheman canadenae 






Me Campyram lineare 
Cornue canadensis 
Epigea repene 
Clintonia borealie 

+ + 
+ + 








Fubus alleghenieneie 
Lycopodium annotinum 
Linnaea borealie 

* I 





Cypripedium acaule 
Lycopodium obscurum 







Lycopodium clavatum 
Commandra umbellata 
Habenaria clarellata 

1 + 




Aralia hispida 
Coptie trifolia 
Aralia nudicaulie 




Oxalia Acetoaella 
Goodyera pubeecene 
Mitchella repene 



Monatropa uni flora 
Lathyrus Taritima 



Epilobiun arguetifolium 
ftumex Acetoeella 



Hudsonia tomentoea 

1 ♦ 


FIGURE 8. Map showing location of releves on the 
Presque Isle tombolo. The size of the plant com- 
munity represented by the releve is indicated by 
the extent of the pattern surrounding the releve 


TABLE 2C. Summary chart of species present in herb 
layer (graminoids and seedlings) and ground layer in 
25 releves on Presque Isle tombolo. 

Relev£ nurter 
total no. of speci es 


poorly drained — * well drained 

2B20 171613 21 U 6 9 24 1816 14 12 10 8 23 *> 19 74 3 1 22 2 
>' 8 20161S 15 1324 814 1411 191623 3022 21 101124 8 9 

Care j- ,:<ligospema 
Cladiurn nariscoides 
Rhynchoqora alta 
Carex rostrata 
Carex muricata 
Eriophorum virginicum 
Schemchzeria palustri* 
Rhynchospora fueca 
Carex Michauxiana 
Carex laeiocarpa 
Carex exilie 
Carex brunnescens 
Unknown #46 
Carex trisperma 
Carex lisnosa 
Scirpue cyperinus 
Glyceria canadensis 
Juncus canadensis 
Carex paupercula 
Juncus pelocarpus 
Carex viridula 
Carex lenticularis 
Unknown #32 
Unknown #42 

Unknown #44 

Unknown #46 

Armophila breviligulata 

Carex arctata 

Carex pensylvanica 

Carex ecoparia 

Unknown #40 

Unknown #41 
Unknown #47 

3 2 2 4 2 2 4 3 

2 2 2 2 

12 2 2 1 

2 2 2 3 


Ptnus Strobus 
Be tula papyri f era 
Picia mariana 
Pinus resinosa 
Acer rub run- 
Abiis balsamea 
Teuga canadensis 
Larix laricina 


1OTI Sp. 

conspicuous mosses 

Cladoma sp. 
conspicuous lichens 

i 5 44 454 4 

2 + 1 
+ 1 

11 +1 


and some 

Chart showing the relationship between topography, vegetation, 
environmental factors . 




Communl ty 

Oomi nant 


Domi nant 
Growth Form 


to Wind 


lake dunes 




gram 1 no Id 



hoi lows 

lake dunes 

Sedge Hoi low 

Cladium/ Juncus 

, : .-. 

grami nold 



Stabl 1 ized 

lake dunes 


Beach Heather 
Pi ne-Bracken 










d me-sic 


Black Spruce 

Picea mariana 






1 Ichen 

Pi ne-Maple 







d mellc 




Teuga/ Betula 







Chamae daphne/ 








f 1 1 led lake 


open bog 



Carex/ Sphagnum 




grami noid 

grami noi d 




Water level is believed to have stabilized, be- 
cause of a halt in erosion at the Port Huron out- 
let between 3200 and 2500 B.P. (Lewis 1969). This 
stage has been designated the Algoma state. Dur- 
ing this period of stability, at 184 m (596 ft) 
MSL, the final barrier beach (S d of Fig. 2 £, 5), 
which forms the northeast shore of the tombolo, 
was probably developed. The subsequent drop to 
the Sault stage, 181 m (586 ft) MSL, undoubtedly 
aided the development of this final Algoma beach, 
exposing more beach material to wind action. Be- 
cause of the short duration of the Sault stage 
and its distance below present lake level, it is 
reasonable to conclude that no shorelines from 
this stage remain visible today. Although the 
present lake level is only about 2 m (80 in) 
below the presumed lake level at the time of the 
tombolo 's inception, the present writer feels 
that dune deposition is responsible for initially 
raising the altitude of the formation sufficiently 
above lake level so that rising water levels of the 
past 2200 years have not totally submerged the 

Additional evidence was sought to verify the 
historical lake-level interpretation of the 
tombolo 's origin. Extensive coring was conducted 
in the large triangular-shaped bog. It was 
ascertained that the bog graded from one meter 
in depth on the edges to eight meters near the 
center. Sediments extracted from the central 
basin contained 1.5 m of lake sediments at the 
base of the core, overlain by 6.5 m of organic 
sediments. Loss of ignition was recorded for 
representative samples taken from the eight meters 
of sediment. Results (Fig. 6) show that the lake 
sediments contain less than 40% organic matter, 
and that the past sediments contain more than 80% 
organics . Lenses of sand were found in many 
cores, possibly the result of overwash during 
high winds, when the enclosing ridges were poorly 
developed. It is concluded that a small tri- 
angular lake existed, enclosed by sand ridges, 
until the lake- level in adjacent Lake Superior 
dropped, causing a comparable drop in level of 
the enclosed lake. At this time the lake may 
have been quite shallow, and the development of a 
bog environment was probably initiated (Fig. 5). 
As the lake level began to rise, due to uplift at 
the St. Mary's River outlet, an increasingly 
deeper bog mat would have developed. 

The history of the bog inferred from the 
sediment stratigraphy, although not conclusive 
evidence, does fit the falling and rising 
sequence of lake level during the past 5500 years. 
Samples 6.54 m below the surface of the bog, near 
the contact of the basal clay with highly organic 
sediment, were dated by radiocarbon analysis as 
4990 + 170 years old (GX-4552) . Lake sediments 
lying below the carbon-dated layer were deposited 
before that in an enclosed o. partially enclosed 
lake formed by the developing spits (S a , S^ of 
Fig. 5). The 1.5 m of lake sediments were 
probably deposited within a relatively short time 
following separation of the small lake from the 
Lake Superior basin. At this time, it is im- 
possible to identify an absolute date for the 
beginning of lake deposition. Evidence supports 
a beginning date that coincides with the ex- 
tension of the spits, that enclosed the lake and 
formed the Presque Isle tombolo. The radio- 
carbon date from the base of the organic sediments, 
and the inclusion of additional time for the 
deposition of lake sediments, puts the inception 
of Presque Isle tombolo within the Nipissing 
time period of 5500 B.P. described by Lewis 
(1969) and Saarnisto (1975). 


The methods utiliz 
tion of plant species 
vegetation assume tha 
distinguished as a di 
According to Mueller- 
a plant community is 
interacting with one 
they modify and depen 
community as defined 
aggregation of organi 
type and having a cha 

ed in mapping the distribu- 

and describing the 
t a plant community can be 
screte unit of vegetation. 
Dumbois and Ellenberg (1974) , 
a combination of plants 
another in an environment 
d upon. Thus, a plant 
for this study refers to an 
sms sharing a common habitat 
racteristic species corn- 

Structural, compositional, and environmental 
criteria were utilized as uniformly as possible 
in identifying and classifying plant communities. 
Boundaries of these communities were located 
primarily by visible differences in structure of 
vegetation and species composition and differences 
in topography. Analysis of aerial photographs 
provide a basis for identification of major 
boundaries, and reconnaissance on foot verified 
the reliability of the boundaries assigned. 
Structural and compositional differences were 
recognized according to height and relative 
importance (percent cover) of the dominant spe- 
cies (Table 1) . Topographic boundaries were 
clearly evident because of difference in elevation, 

Secondarily, environmental criteria, such as 
exposure to wind and moisture condition, were 
estimated. Exposure was considered on a scale 
from low (protected) to extreme (unprotected) . 
Moisture conditions were estimated by combining 
information on exposure, height above water 
table, and known ecological requirements of spe- 
cies present within the habitat. Single plots 
in each unit were analyzed according to the 
Braun-Blanquet releve method. The releve' plot 
was large in relation to relief so that slope 
and aspect were of lesser significance than 
height above water table and exposure to wind 
action. Each releve was rated on a scale from 
hydric to xeric according to status of available 
moisture . 

The twenty-five 
arranged according 
species (Table 2) . 
carried out by vis 
lists, placing tog 
greatest number of 
releves can be div 
major groups: hydr 
compilation of dat 
positive correlati 
available moisture 
species align runs 
left of the tables 
present in hydric 
some cases sharply 
xeric conditions p 

releves were arranged and re- 

to presence or absence of the 

This is a numerical method, 

ual inspection of the species 

ether those releves with the 
species in common. The 

ided rather readily into three 

ic, mesic, and xeric. The 

a in this manner show a 

on between species presence and 
The gradient along which 
consistently from the top 
to the bottom right. Species 

communities gradually or in 
drop in importance as more 

revail . 

The pattern is consistently visible in all 
canopy layers: trees, saplings, seedlings, shrubs, 
and herbs. As an example, in the shrub I layer, 
(Table 2) Chamaedaphne calyculata , Myrica gale, 
Andromeda gl aucophy 1 1 a , and Kalmia polio folia 
dominate the hydric communities but rarely occur 
in the xeric communities . As those species drop 
in importance, Ledum groenlandicum , Gaylussaccia 
baccata , Vaccinium angusti folium, and Vaccini urn 
myrtilloides appear as dominants in the mesic 
communities . The trend continues with different 
species present and dominant in the xeric 
communities: Juniperus communis , Spirea alba, 
Prunus pumila, and Arctostaphylos uva-ursi. 


A positive correlation also exists between 
topography and the mosaic of plant communities 
found on Presque Isle tombolo. Similar patterns 
have been established in coastal systems by 
Oosting and Billings (1942), Martin (1959), 
and Au (1974). Martin's classic work on the 
vegetation of Island Beach State Park, New 
Jersey, is a comprehensive study of the inter- 
relationships among vegetation, topography, and 
environment in a marine coastal dune system. 
On Presque Isle tombolo the same interrelation- 
ships exist on an inland-lake coastal-dune 
system, where salt spray is not a factor. The 
conclusions reached in both studies are striking 
in their similarities. In both cases the exist- 
ing zoned mosaic of vegetation corresponds with 
the zoned mosaic of topography. The major 
difference in the two systems is the salt versus 
fresh-water influence. Species composition and 
distribution correlates with species level of 
salt tolerance in the marine dune system. This 
factor, though absent in the inland-lake dune 
system, does not alter the similarity in distri- 
bution of life-form types. The similar patterns 
found in both marine and inland-lake dune systems 
demonstrate the existence of an analogous 
combination of environmental and physical 
processes . 

According to Martin (19 
a central and diagnostic p 
of a complex environment, 
intensity of limiting fact 
upon the plants directly, 
an indicator of plant dist 
type on the Presque Isle t 
topographic zones exist (F 
interdune hollow, stabiliz 
and inter-ridge trough or 
The zones vary significant 
position of exposure, and 
relief of the tombolo is n 
seven meters . 

59) , topography occupies 
osition in the analysis 
because it controls the 
ors but does not act 

Topography serves as 
ribution and habitat 
ombolo. Four distinct 
ig. 7): active dune, 
ed dune/beach ridge, 
filled lake basin, 
ly in configuration, 
elevation, even though 
ever greater than 

The active dunes are low (Releve #2) combin- 
ing with the beach to form the mid-section of the 
northeast shoreline of the tombolo. The low 
relief of this section, directly exposed to 
northeast winds, is maintained by overwash during 
major storms (Godfrey 1974) . The interdune 
hollow zone (Releve #24) , is located in this same 
area in wet depressions behind and between the 
active dunes. The stabilized dunes can be 
separated into four distinct regions by location 
on the tombolo and by vegetation cover. (1) One 
area lies south of and adjacent to the active 
dunes, bordering the northeast shore (Releve #1). 
A distinct rise in elevation separates these 
stabilized dunes from the overwash active-dune 
area. The greater deposit of sand in this area 
is explained by its position on the tombolo in 
relation to longshore currents and angle of ex- 
posure to wave and wind actions . These dunes are 
stabilized by a xeric population of low 
herbaceous and shrubby plants. (2) A ridge of 
stabilized dunes (Releve #3) lies directly behind 
most of the active dunes, except where they have 
been obliterated by overwash. It is a narrow 
ridge stabilized predominantly by the ericaceous 
shrub Chamaedaphne calyculata (Leather leaf) and 
by a single line of scattered trees. It is 
bordered on its back side by open water of the 
flooded lagoons. (3) The third area (Releves 
#12, 14, 16 & 18) is much like the second in 
vegetation cover but distinct in form. These 
ridges are small, parallel, sand ridges (4 to 
10 m wide) exhibiting no dune development. 

(4) The fourth area of stabilized dunes, adjacent 
to and southwest of the parallel sand ridges, are 
the broad "Nipissing" dune ridges (Relev6s #5 & 
7) , forested by confier and mixed northern hard- 
woods . The inter-ridge troughs (Releves #11, 
13, 15, 17, 21) 5-10 m strips of sphagnum bog, 
separate the parallel sand ridges . The filled 
lake basin (Releve #20) , now a 16 ha triangular- 
shaped bog, is enclosed by the "Nipissing" dune 
ridges . 

The relationship between topography and plant- 
community distribution is demonstrated on a pro- 
file (Fig. 7, located on Fig. 8 B-B'). Dominant 
species present along the transect are 
symbolically illustrated on the topographic pro- 
file so that representative communities of most 
habitats were included within the transect. The 
profile clearly demonstrates that the occurrence 
of a specific vegetation type is associated with 
a given topographic unit. Tabel 3, compiled 
with information from both the profile and releve 
data, defines twelve distinct communities, each 
characterized by a unique combination of 
physiographic parameters and dominant vegetation. 

Successional Relationships . -- In the absence 
of longterm data on successional patterns on 
Presque Isle tombolo, clues to the probable 
successional relationships must be gathered from 
the existing vegetation. On a broad scale, 
succession on the tombolo does not correspond 
with the age of the substrate. Trees of the 
oldest age class (approximately 100 years old) 
presently exist on the oldest "Nipissing" dune 
ridges and also on the youngest "Algoma" dune 
ridge. Thus, the existing vegetation cover does 
not indicate the age of the landform but is a 
product of dynamic environmental, biotic, and 
physical processes. 

It appe 
level. As 
exposed fo 
newly expo 
that previ 
manner, ac 
dunes , and 
sion (dune 
to black o 
evolve . 

ars that 
ly depen 

water 1 
r coloni 
sed surf 
ously bo 
tive dun 

a tradi 

dent on the 
evel drops, 
zation by p 
aces provid 
rdered the 
es can deve 
tional patt 
s to jack p 

oak, Olson 

on the tombolo is 
status of water 
new surfaces are 
ioneer species . The 
e protection to areas 
lake edge. In this 
lop into stabilized 
em of dune succes- 
ine , or white pine 
1958a) might 

However, the trend of 
the past 2000 years cause 
traditional patterns of s 
evident on the southwest 
dunes . Exposure to wind 
localized areas created s 
(populated with species c 
active dunes) , midst the 
The dead trees on the out 
another example. The flo 
by the rising water table 
establishment of species 
hydric environment. 

rising water levels over 
s a reversal from 
uccession. This is 
edge of the forested 
and wave action has in 
mall mobile dunes 
haracteristic of the 
stable forested dunes, 
er beach ridges are 
oding condition caused 

requires the 
more tolerant of a 

The dynami 
smaller scale 
graphic zones 
stable within 
Ammophi la bre 
destruction o 
(Beach Heathe 

cs of equilib 
development c 
in each of t 
The active 
a cycle of s 
by pioneer s 
v i 1 i gul a ta ) , 
f the dunes . 
r community) 
stage between 

rium and patterns of 
an be followed on a 
he different topo- 

dunes are currently 
and deposition, 
pecies (predominantly 
and overwash and 

The stabilized dunes 
are potentially an 

active dunes and 


forested dunes. The establishment of equilibrium 
on these dunes is maintained by the interaction 
between vegetation involved in stabilization and 
physical forces, i.e., wind action (blowouts), 
active in the destruction of the stability. 

The forested dune ridges appear to be quite 
stable, dominated by species capable of re- 
producing beneath their own kind. Olson's study 
of rates of succession and soil change on the 
Lake Michigan sand dunes shows that soil maturation 
occurs within the first 1000 years after 
stabilization. He found that little improvement 
in soil fertility occurs after this length of 
time. The condition of low fertility favors 
vegetation with low nutrient requirements . In 
this environment, there is little likelihood of 
improving nutrient conditions as these species are 
unable to return nutrients to the soil. Thus, 
except in isolated locations, according to 
Olson (1958a), it is unlikely that succession on 
dune surfaces reaches the climax mesophytic 
forest that Cowles (1899) spoke of in his model 
of sand-dune succession in the Lake Michigan dunes. 
The dune-ridge forests of Presque Isle tombolo are 
predominantly forests composed of species with low 
nutrient requirements, Pinus resinosa , Pteri.dium 
aquilinum , and Vaccinium angusti folium. The fire 
history, evidenced by fire scars and historical 
record, is an additional factor determining the 
dominance and continuance of the Pine forest on 
the dune ridges . 

Successional development on Presque Isle tombolo 
is controlled by physical and environmental forces 
unrelated to the age of the substrate. The 
existence of active dunes, dunes stabilized by low 
shrubby growth, and dunes stabilized by forest 
growth does not necessarily represent serai 
stages in autogenic succession. Succession is 
primarily an interzonal phenomena, characterized 
by cyclical rather than linear patterns of 


The complex pattern of plant 
in such close proximity, is the 
speculation. What was the geol 
gave rise to such a diverse Ian 
have climate, winds, currents, 
the development of the existing 
Plant zonation functions as a k 
the subtle continuum of change, 
into troughs, and troughs into 
interrelationship of biotic and 
involved in the development of 
tombolo demonstrates the divers 
vegetation and topography which 

communities, with- 

object of 
ogic history that 
dform? What effects 
and storms had on 

een indicator of 

as ridges grade 
ridges . The 

physical factors 
Presque Isle 
ity found in 

are dynamically 

According to 
any historical 
abundance of th 
data can be int 
forms or physic 
stand the natur 
process . " The 
geomorphic and 
of existing top 
and the results 
analyses, provi 
of an hypothesi 

Wright (1966) : "The accuracy of 
reconstruction depends on the 
e primary data, and on how well 
erpreted. If the record is land- 
al sediments, one should under- 
e of the geomorphic and geologic 
synthesis of information from 
geologic history, interpretation 
ographic features on the tombolo, 

from sediment and radiocarbon 
de the basis for the development 
s . 

high period of the Nipissing lake stage, 188 m 
(608 ft) MSL. A series of parallel ridges, 
culminating with a final barrier beach formed on 
the northeast shore of the tombolo, are evidence 
of the continued retreat of the lake level. The 
outer beach is identified as an Algoma beach, 
formed between 3200 and 2500 B.P. History of the 
bog (B of Fig. 2) , inferred from the sediment 
analyses, fits the falling and rising sequence of 
lake levels in the lake Superior basin during the 
past 5500 years. A radiocarbon date 4990 + 170 
years B.P., from the deepest organic sediments 
in this bog and the conclusions from the sediment 
analyses together support the lake-level inter- 
pretation of the origin of Presque Isle tombolo. 

Billings (1952) acknowledges th 
complexity of the interrelationshi 
plant and its environment is almos 
discourage any attempts at a compl 
and synthesis '. In the study of P 
tombolo, a preliminary attempt has 
to interpret the floristic pattern 
the geomorphic processes active in 
of the existing topography. A tot 
ing of the ecological processes in 
the Presque Isle tombolo awaits mo 
investigation of additional biotic 
parameters . 

at "the 

ps between the 
t enough to 
ete analysis 
resque Isle 

been made 
s in terms of 

the formation 
al understand- 
teracting on 
re specific 

and abiotic 

The origin of the ridges and troughs that form 
Presque Isle tombolo are related to water-level 
fluctuations, wind, and currents. The inception 
of the tombolo began approximately 5500 years 
ago, as the lake level began to retreat from the 

Boundaries between adjacent plant communities 
are visibly distinct. When organized into the 
summary table (Tabel 2) , certain species are 
found to possess sufficient-physiologic 
adaptability to enable them to occur in more than 
one community, so that a nearly continuous 
gradient seems to exist from hydric to xeric. 
By measure of structure and composition, five 
different vegetation types are recognized within 
this gradient: lake dunes, northern conifer 
forest, northern mixed forest, shrub carr, and 
open bog. Four different topographic zones are 
identified: active dunes, interdune hollows, 
stabilized dune/beach ridges, and inter-ridge 
troughs or filled lake basin. A combination of 
the following parameters: topographic zone, 
vegetation type, moisture status, degree of 
exposure to wind action, and dominant species 
define twelve separate plant communities (Table 

It is difficult to evaluate accurately the 
relative importance of the parameters studied. 
However, Presque Isle tombolo clearly demon- 
strates the interrelationship and interdependence 
of topography, environment, and vegetation. A 
rich flora exists because of the abundance of 
diverse habitats. Any number of variables com- 
bine to define the uniqueness of a given site. 
One or more variables may express their 
dominance at a given location, for example: 
extreme wind conditions (a limiting factor to 
the stature and available moisture for vegetation 
in the Beach Heather community) , exposure to wave 
action and over-wash (a limiting factor to the 
development of dune vegetation) , nutrient-poor 
soil (a limiting factor to species composition 
in forest flora), moisture conditions (a 
limiting factor in bog development) , and sand- 
stabilizing vegetation (a limiting factor to 
dune mobility) . 

The interwoven relationship of geomorphic 
processes and topography have created a landform 
that combines with environmental factors in many 
distinct combinations. The habitat that results 
requires plants capable of adapting and co- 
existing in the ecosystem they comprise. The 
environmental factors, the topography, and the 


geomorphic processes are reciprocally modified 
by the vegetation existing in the habitats. The 
Presque Isle tombolo emerges as a dynamic system 
demonstrating the vital interrelationships among 
physical and biotic processes. 


ANONYMOUS. 1974. Stockton Island Survey: An 
ecological survey and environmental impact 
study of Stockton Island, Apostle Islands Na- 
tional Lakeshore. Northland College, Ashland, 
Wise. 71 pp. 
AU, SHU-FUN. 1974. Vegetation and Ecological 
processes on Shackleford Bank, North Carolina. 
Nat. Park Service Scientific Monograph Series. 
No. 6. 8 6 pp. 
BEALS, E. W. and G. COTTAM. 1960. The forest 
vegetation of the Apostle Island, Wisconsin. 
Ecology. 41(4) :743-751. 
BILLINGS, W. D. 1952. The environmental complex 
in relation to plant growth and distribution. 
Quart. Rev. Biol. 27:251-165. 
COLLIE, GEORGE. 1901. Wisconsin shore of Lake 

Superior. Geol. Soc . Amer . Bull. 12:197-216. 
CONWAY, V. M. 1949. The bogs of Central Minne- 
sota. Ecol. Monogr. 19:173-206. 
COOPER, W. S. 1958. Coastal Sand Dunes of 

Oregon and Washington. Geological Society of 
America. Memoir 72. 169 pp. 

1967. Coastal Dunes of California. 
Geological Society of America. Memoir 104. 
131 pp. 
COWLES, H. C. 1899. The ecological relations ot 
the vegetation on the sand dunes of Lake Michi- 
gan. Bot. Gas 27:95-117, 167-202, 281-308, 
CURTIS, J. T. 1959. The vegetation of Wisconsin. 
An ordination of plant communities. Univ. 
Wisconsin Press, Madison, Wise. 657 pp. 
CUSHING, E. J. and H. E. WRIGHT, JR. 1965. 

Piston corers for lake sediments. Ecology 46: 
ENGSTROM, W. 1972. Spatial patterns in beach 
morphology and sedimentation in the Apostle 
Islands of northern Wisconsin. Ph.D. Dissert. 
Univ. of Wise, Madison. 236 pp. 
EVANS, 0. F. 1972. The Origin of Spits, Bars, 
and Related Structures. p. 52-72. in: Spits 
and Bars. Schwartz, M. L. (ed.), Dowden, 
Hutchinson, and Ross. Stroudsburg, Penn. 
4 52 pp. 
FARQUHAR, O. C. 1972. Stages in Island Linking, 
p. 307-330. In: Spits and Bars. Schwartz, 
M. L. (ed.), Dowden, Hutchinson, & Ross. 
Stroudsburg, Penn. 452 pp. 
FARRAND, W. R. 1960. Former shorelines in west- 
ern and northern Lake Superior Basin. Ph.D. 
Dissert. Univ. of Michigan, Ann Arbor. 226 pp. 

. 1969. The quaternary history of 

Lake Superior. 12th Conf. Great Lakes Research 
Proceedings: 181-197. 
GATES, F. C. 194 2. The bogs of Northern lower 

Michigan. Ecol. Monogr. 12:213-254. 
GLEASON, H. A. and A. CRONQUIST. 1963. Manual 
of vasuclar plants of northeastern United 
United States and adjacent Canada. D. Van 
Nostrand Co., New York. 810 pp. 
GODFREY, P. J. and M. M. GODFREY. 1974. The 
role of overwash and inlet dynamics in the 
formation of salt marshes on North Carolina 
Barrier Islands. P. 407-428. in. Ecology of 
Halophytes. Reinold, R. J. and W. H. Queen, 
(eds.), Academic Press, New York. 605 pp. 
GULLIVER, F. P. 1899. Shoreline topography. 

Proc. Amer. Acad. Arts and Sciences. 34:151- 

HEINSELMAN, M. L. 1970. Landscape evolution, 

peatland types, and the environment in the 

Lake Agassiz peatlands natural area, Minnesota. 

Ecol. Monogr. 40:235-261. 
HOUGH, J. L. 1958. Geology of the Great Lakes. 

University of Illinois Press, Urbana, Illinois. 

313 pp. 
JOHNSON, D. W. 1919. Shore processes and shore- 
line development. Wiley, New York. 524 pp. 
KING, C. A. M. 1959. Beaches and Coasts. 

.. Edward Arnold, Ltd., London. 403 pp. 
KUCHLER, A. W. 1964. Potential Natural Vegeta- 
tion of the conterminous United States. 

American Geographic Society, New York. 116 pp. 
LAING, C. C. 1958. Studies in the ecology of 

Ammophila brevi ligulata L. seedling survival 

and its relation to population increase and 

dispersal. Bot. Gaz. 119:208-216. 
LEVERETT, FRANK. 1929. Moraines and Shorelines 

of the Lake Superior Basin. U.S. Geol. Survey 

Prof. Paper. 154-A:l-72. 
LEWIS, C. F. M. 1969. Late Quaternary history of 

lake levels in the Huron and Erie basins. 

Proc. 12th Conf. Great Lakes Res. Int. Assoc: 

LEWIS, W. V. 1931. The effect of wave incidence 

on the configuration of a single beach. Geog. 

Jour. 78:129-143. 
. 1932. The formation of Dungeness 

Foreland. Geog. Jour. 80:309-325. 
LOY, W. G. 1962. The Coastal geomorphology of 

western Lake Superior. Univ. of Chicago M.S. 

thesis. Chicago, Illinois. 85 pp. 
MARSHALL, ERNEST. 1967. Lake Superior ice 

characteristics. 10th Conf. Great Lakes 

Research Proceedings : 200-214 . 
MARTIN, W. E. 1959. The vegetation of Island 

Beach State Park, New Jersey. Ecol. Monogr. 

29(1) :l-46. 

Aims and Methods of Vegetation Ecology. John 

Wiley & Sons. New York. 547 pp. 
OLSON, J. S. 1958a. Rates of succession and soil 

changes on southern Lake Michigan Sand Dunes. 

Bot. Gaz. 119(3) :125-169. 
. 1958b. Lake Michigan dune develop- 
ment. I. Wind velocity profiles. Jour. Geol. 

. 1958c. Lake Michigan dune develop- 

ment. II. Plants as agents and tools of geo- 
morphology. Jour. Geol. 66:345-351. 

1958d. Lake Michigan dune develop- 

ment. III. Lake-level, beach, and dune oscil- 
lation. Jour. Geol. 66:473-483. 

OOSTING, H. J. and W. D. BILLINGS. 1942. Fac- 
tors effecting vegetational zonation on coast- 
al dunes. Ecology 23:131-142. 

SAARNISTO, MATTI. 1975. Stratigraphical studies 
on the shoreline displacement of Lake Superior. 

SCOTT, IRVING. 1926. Ice push on lake shores. 
Mich. Acad. Sci. Arts. Letters Papers. 7:107- 

SHARP, R. P. 1953. Shorelines of the glacial 
Great Lakes in Cook County, Minnesota. Amer. 
Jour, of Science. 251:109-139. 

TANS, W. E. and R. H. READ. 1975. Recent Wiscon- 
sinrecords for some interesting vascular plants 
in the western Great Lakes Region. The Michi- 
gan Botanis. 14:131-143. 

TAYLOR, F. B. 1894. A reconnaissance of the 
abandoned shorelines of the south coast of 
Lake Superior. Amer. Geol. 8:365-383. 

THWAITES, FREDIRK. 1912. Sandstones of the 

Wisconsin coast of Lake Superior. Wis. Geol. 
and Natural Hist. Survey Bull. No. 25. 117pp. 


TURNOCK, W. and D. B. LAWRENCE. 1953. Measure- 
ment of the level of the ground water at the 
Cedar Creek Natural History Area, central 
Minnesota. University of Minnesota, Dept. of 
Botany Report. 11 pp. 

USDC. 1974a. Wisconsin, p. 450-452. in: Cli- 
mates of the states. Vol. I., Water Info. 
Center, Inc., Washington, D.C., 480 pp. 

USDC. 1974b. Duluth, Minnesota. Local Climato- 
logical Data. Annual Summary with Comparative 
Data. National Oceanic and Atmospher. Admin. 
Environm. Data Serv. Nat. Climatic Center, 
Asheville, N.C. 

USDC. 1976. Great Lakes Pilot. Nat. Oceanic 

and Atmospheric Adm., Washington, D.C. 637 pp. 

WRIGHT, H. E., JR., 1966. Stratigraphy of lake 
sediments and the precision of the paleocli- 
matic record. p. 157-172. in: Proceedings 
of the International Symposium on World Cli- 
mate. Royal Meteorological Society, London. 
324 pp. 

CUSHING. 1965. Coring devices for lake 
sediments. p. 494-520. in: Handbook of 
paleontological techniques. Kummel, B., and 
D. M. Raup, (eds.) , W. H. Freeman Co., San 
Francisco. 852 pp. 

WRIGHT, H. E., JR. and W. A. WATTS. (with 
contributions by S. Jerlgersma, J. C. B. 
Waddington, T. C. Winter, and J. Ogawa). 1969, 
Glacial and vegetational history of north- 
eastern Minnesota. Minnesota Geol. Surv. 
Spec. Publ. 11:1-59. 




Donald J. McGraw 


Nearly one-half a millenium after Christopher 
Columbus first landed in the New World, the citi- 
zens of America made preparations to celebrate the 
four hundredth birthday of the discovery. It was 
the intention of ardent men across the land to 
assemble a great exposition to demonstrate the 
achievements of America. The year of the great 
fair was to be 1891-92. As is often the case with 
such gargantuan adventures, the sheer weight of 
the machinery bogged down into the morass of good 
intentions and it was not until October 1892 that 
Chicago hosted the World's Columbian Exposition, 
the "Chicago World's Fair." 

A call went out acros 
fair committee to states 
to prepare exhibitions o 
ter demonstrate to the wo 
dustrial, and cultural a 
In a few counties, and i 
response to Chicago was 
Tulare in east-central C 
a section of Giant Sequo 
teum (Lindl.) Buchholz, 

s the country 
, regions, an 
f local succe 
rid the agric 
chievements o 
n one in part 
unique. The 
alifornia cho 
ia, Sequoiade 
1939, as its 

from the 
d counties 
sses to bet- 
ultural, in- 
f America, 
icular, the 
County of 
se to send 
ndron gigan- 


on gigi- 
ibit. 3 

It was not the first time such an exhibit trav- 
eled east from California. Indeed, the matter of 
convincing fair-goers that such tree sections of 
the immensity common to the Giant Sequoia were 
real was no mean task. At the centennial fair of 
America's founding in Philadelphia in 1875, a 
sequoia section so shoddily cut and reassembled 
as to be unconvincing was labeled a "California 
Hoax." That demeaning appellation did not, how- 
ever, fall upon the Tulare exhibit 18 years later 
in Chicago. It is one of the threefold purposes 
of this paper to indicate why the American populace 
accorded these two apparently similar exhibits with 
such differing receptions. 

The story of the 189 3 Tulare section has been 
recounted upon several occasions in rather dis- 
parate publications, but in each only highlights 
have been given and often only a modicum of his- 
torical accuracy is present. It is the second and 
main purpose of this study to present a fuller and, 
hopefully, more accurate history of the subject 
and, in so doing, to elaborate further upon already 
published, but abbreviated, accounts elsewhere. A 
variety of previously unpublished materials has 
come to light and assists in producing a fuller 

The cutting operation of the tree was brilliant- 
ly recorded by a pioneer photographer, C. C. Curtis, 
and is a vivid account of the methods of prepara- 
tion of the exhibit. Although a number of the 
scenes of the operation have been published, others 
have never been made available. It is the third 
purpose of the current paper to make the photo- 
graphic history available ir print. 

The research on this subject was begun in 1970 
during the author's first summer season as an inter- 
preter on the staff of the General Grant Grove area 
of Sequoia and Kings Canyon National Parks. 


In historic Kern County, California, efforts 

had for some time been underway to select exhibi- 
tion material for the Exposition. In a plan form- 
ulated in meetings between Kern and Fresno counties, 
it had been decided to ask Tulare County to join in 
a three-county central California exhibit con- 
sortium. A communication sent to the interest- 
ed commission already extant in Visalia, Tulare 
County, was received favorably and its secretary 
responded that it would accept the tri-county 
invitation. ^ The commission then quickly approach- 
ed the County Board of Supervisors to request an 
appropriation of $5000 in order to prepare a fit- 
ting exhibit. The Board took the matter under 
advisement with a promise to respond within a day 
or two. Apparently that did not happen, for on 
June 11 the visalia Daily Times indicated that 
it had reminded the Board and others who read the 
newspaper that "the matter [of the appropriation] 
must not be dropped and all persons who have the 
welfare of the county at heart should give the mat- 
ter their aid and attention." 5 The editors went 
on to suggest a petition be circulated to urge 
action on the part of the Board. 

Several petitions wer 
sition to the appropriat 
remarked that the matter 
itself down to a one-sid 
explicably the matter wa 
and early August when a 
articles, not only in th 
Central Valley (Californ 
sions on the Giant Sequo 
for the exhibit. Other 
not $5000 but $15,000 ac 
We are left with a gap i 
which would enlighten us 
reversal took place. 

e circulated — all in oppo- 
ion. The Times editors 

"seems to have settled 
ed question." 6 Yet, in- 
s reversed by late July 
long series of newspaper 
e Times, but in other 
ia) dailies, began discus- 
ia that had been selected 
evidence indicates that 
tually was appropriated, 
n the historical record 

on how this interesting 

The Times first introduced to the county resi- 
dents a tree cut on the land of Mrs. M. C. K. 
Shuey in Tulare County which it claimed had been 
selected as the exhibit. In fact, this tree was 
apparently never intended for the Columbian Exposi- 
tion. The most authoritative discussion of all 
Sequoiadendron giganteum trees cut for any exhi- 
bition purposes at any time occurs in the 
Historical Bulletin of the Tulare County Histori- 
cal Society. No explanation of the "Shuey tree" 
has been forthcoming regarding its final disposi- 
tion. The Times went on, however, some day later 
to provide correct information on the actual exhi- 
bit tree, the General Noble. 

The Noble Tree was selected as the one to repre- 
sent Tulare County in Chicago. It is an interest- 
ing aside that the tree itself was not actually 
within the confines of the county boundary lines 
of Tulare, but slightly to the north in that of 
Fresno. Evidently this never led to any acrimony 
or disagreements. 

The name General Noble, li 
many other famous Giant Sequo 
military leader from the peri 
the States (but see below) 
ently the namesake of John Wi 
soldier, lawyer, and finally 
Interior. A native of Lancas 
no evidence he ever saw the t 
virons . Whether he ever saw 
its existence as an exhibit a 

ke the names of so 
ias, recognized a 
od of the War Between 
The Noble was appar- 
llock Noble (1831-1912), 
Secretary of the 
ter, Ohio, there is 
ree in its native en- 
it or even knew of 
t the fair is equally 


unknown. However, the exhibit 
most prominent place in the mo 
at the fair, the United States 
Therefore, it would not seem s 
that Noble saw it, or was, at 
section. The official history 
calls the tree section the Joh 
and it is a fact that Noble, a 
strongly supported the America 
tion in regard to land law rev 
Who named the tree for him and 
if we depend upon Gray's small 
sequoias . 13 

section held the 
st notable building 

Government Building, 
urprising to surmise 
least, aware of the 

written on the fair 
n W. Noble tree, 
s Interior Secretary, 
n Forestry Associa- 
isions in 1891.12 

when is not known, 

volume on named 

Most trees named for 
men who fought under th 
the States and who late 
adventure and fortune i 
bering enterprises. Bu 
eral miles of the Noble 
in the 1860's to 1880's 
pression that the Noble 
shortly before the cutt 
it was not by appreciat 
cal lumbermen appreciat 
favorable forestry stan 

generals were so named by 
em during the War Between 
r came west often seeking 
n the Sierra Nevadan lum- 
t no other tree within sev- 

was named after a general 
It is the author's im- 

was so christened only 
ing. It is believed that 
ive war veterans but by lo- 
ive of Secretary Noble's 

The story of the Noble really begins with its 
cutting and it is that period about which most is 


On August 12, 1892, the first of three cross- 
cuts was made into the General Noble and its ex- 
istence as an exhibit for the next 39 years was 
established. I 4 Its existence as a living thing, 
however, came to a halt. 

An anonymous author in the Historical Bulletin 
of the Tulare County Historical Society said that 

At first they [exhibits of sequoia] 
were greeted with skepticism but 
gradual ly the reports of eyewi tnesses 
and the general distribution of photo- 
graphs overcame doubts and California's 

Big Trees were accepted as facts. 

1 5 

It is true than many more people had seen the 
great tree in situ by the 1890 's than by 1876 when 
the centennial "Hoax" was seen. Further, the photo- 
graphic evidence provided by Curtis was undeniable. 
But a third factor was undoubtedly responsible for 
the belief that the exhibit of 1893 was not a hoax. 
The reason was that the Noble section was very 
carefully prepared as compared to the Centennial 
Tree. I 6 The significant role of photographic evi- 
dence is considered in the next section, but the 
actual cutting operation must first be detailed 

The General Noble grew on a piece of land just 
about three airline miles north of what was then 
General Grant National Park in a magnificent 
Sequoia grove known as the Converse Basin. The 
land was owned by the U.S. Government and was a 
forest reserve. A contract for sequoia cutting 
operations was held for that area by the Smith and 
Moore Company, and hence, they owned the Noble. 

The Smith an 
thousands of ac 
porated. Austi 
dent, was an ex 
home in San Fra 
Hiram C. Smith, 
the woods boss 
Kings River ope 
Basin. I 7 

d Moore Company had purchased 
res of land and, by 1888, had incor- 
n D. Moore, the company's presi- 
travagant character with a showplace 
ncisco. The more practical partner, 

the vice president, was actually 
and assumed direct charge of the 
ration which included the Converse 

The World's Fair commission of Visalia, with 
its $15,000 in hand, purchased the Noble (for 

$5,000) from Smith and Moore, who opened contract 
bids to cut and haul the sections. The success- 
ful bidder was a Mr. Burr Mitchell of the nearby 
village of Miramonte. 

Very few roads existed, even by 1892, in that 
heavily lumbered basin. For that reason, a 
contract to build a wagon road to haul out the 
sections was granted Mr. R. Ball of LeGrand, 
California, by Smith and Moore. 19 The road cost 
$1,500. 20 When the road was completed, it was pos- 
sible to haul the sections to the valley hamlet 
of Monson, which was the railhead. 

The first cut made that mid-August was 52 feet 
above the level of the ground, according to most 
reports. That figure is probably correct in 
that a surveyor, Fred Beraria, had made extensive 
studies of the tree. 22 The diameter, at 52 feet, 
was 19 feet 6 inches, according to Beraria. 

Although sequoias are routinely cut some dis- 
tance above the ground to avoid the so-called 
"butt swell" at their bases, the matter of 52 
feet might seem excessive. The reason for choosing 
that height was to allow for considerable quality 
material for exhibition below that level and above 
the swell. 

Although a number of men were present during 
the cutting (they are named in the legend to Fig. 
1), only five were actually involved in felling 
the tree. Gray notes that Burr Mitchell, Will 
Irwin, Dayton Dickey, and Jesse Pattee were in- 
volved. But Pattee (d. 1951) , about whom most is 
known, has indicated that "Captain" Jamison 23 was 
the foreman. 2 The story of the cutting, which 
took 13 days, has been given in uneven sketches 
in Gray, the Historical Bulletin and in a variety 
of contemporary newspaper articles, as well as 
some more recent publications. 25 Not previously 
available in its entirety, however, is the text 
of a wire-recorded interview of Jesse Pattee done 
some years ago. 2 '* A similar interview with C. C. 
Curtis is also enlightening. Blending these 
sources, the story becomes more vivid, but no less 

Jamison himself did not actually handle any 
cutting tools; only Pattee and his colleagues did. 
"Two Scotsmen, riggers from San Francisco, helped 
with scaffolds and ropework," Pattee remembered. 
Their names are unrecorded, but they may be among 
those pictured in Figure 1. As the main portion 
of the tree toppled from the first cut at 52 feet, 
it was expected to clear the tall stump with no 
difficulties. Indeed, a considerable number of 
falling wedges were inserted (by Mitchell) in the 
saw cut to assure a clean fall. But such a fall 
did not occur. The massive bulk of the tree slip- 
ped back upon the stump and smashed the scaffold- 
ing, leaving no escape route for Pattee and his 
partner at that dangerous moment. As the tree 
fell, the two men jumped onto the tall stump as 
their only avenue of escape. They lay there, 
face down, arms and legs spread, for a period of 
20 minutes. It took that long for vibrations to 
cease before they could stand upright.' The un- 
even fall broke off a large portion of bark and 
sapwood from the intended exhibit material below 
the cut. Curtis recorded on film that splintered 
piece, but the photograph is presumed no longer 
extant. He noted that the crew decided to "patch" 
on the splinter and said, "No one seemed to notice 
[the patch] at Chicago or anywhere else." The 
repair must be laid as a tribute to the lumber- 
jacks, for it was just such patching of even the 
properly cut parts on the 1875 tree that were so 
shoddy as to leave the impression of a hoax. 

The next step was to remove staves of bark and 
sapwood each 14 feet in length The interior 


(heartwood) of the stump was hollowed out and, ac- 
cording to Pattee, discarded. Figures 8-11 show 
this very well. Pattee indicated that the sec- 
tions were marked, not with paint or something sim- 
ilar, but by a coded (?) pattern of nails driven 
into each section. That facilitated reassembly in 
Chicago, which was not done with any of the cut- 
ting crew present so far as can be determined. 

After the staves were removed, a fresh stump 
surface 38 feet above the ground existed. A 2-foot- 
thick wheel section was removed in pie-wedge sec- 
tions. This became the floor for the upper "room" 
of the final two-story high exhibit. Another 14 
feet of staves were removed. This yielded two 
"rooms," one above the other and each with 14-foot 
high walls. The upper "room" had a 2-foot thick 
floor. The lower "room," then, had a ceiling, but 
the upper "room" was without one. 

Ball (builder of the wagon road) hauled the care- 
fully labeled and crated sections from the Converse 
Basin to the rail line at Monson. There the boxes 
were shipped to Chicago. No shipping date is known, 
but it must have been about August 25 or within a 
few days after that. 

The entire operation was recorded masterfully by 
Curtis, whose story in regard to this entire project 
is a poignant one. 

Charles C. Curtis and His Photographs 

A native of Marshalltown, Iowa, Curtis (dates be- 
lieved to be 1862-1955) first came to California in 


He joined a studio in Hanford, California, 

in 1884 with a partner named Tandy who had adver- 
tised for an apprentice to learn the photographic 
trade. Shortly thereafter, he joined the now-fa- 
mous Kaweah Colony, one of several well-known ex- 
periments in communism in the United States during 
the 19th century. Headquartered at Visalia, the 
colony sought to cut timber, including Giant Se- 
quoias, in the area of Giant Forest in what soon 
came to be known as Sequoia National Park. The 
federal proclamation creating the Park in 1890 fore- 
stalled the cutting plans. The Colony dissolved 
not long afterward. 3D 

Thereafter, all 11 summer seasons, Curtis did 
photographic work in the sequoia forests of the 
High Sierra in and near Kings Canyon and Sequoia 
National Parks. Additionally, he engaged in such 
work in a variety of nearby central valley com- 
munities . 

Just prior to the cutting of the Noble, Curtis 
made an agreement with Smith and Moore to photo- 
graph the entire process. It was "no firm deal," 
he said. 3 1 Nevertheless, the loose contract en- 
sured him funds to develop and print "thousands of 
[photographic] prints." 

Upon completion of the field work and prepara- 
tion of prints, he went to the fairgrounds in 
Chicago with the intention of selling copies to 
fair visitors. His discovery that only one Chicago- 
based firm (The Werner Company) had sole photograph- 
ic concession rights came as a rude shock. It was 
possible, though, to obtain sales-licensing from 
that company for $250 per sales booth and 25% of 
the gross receipt. The amount was beyond his 
means and so he telegraphed Smith and Moore at 
their corporate headquarters in San Francisco. He 
received no reply for several days. In a Chicago 
newspaper he recalled reading the saddening news 
that Smith and Moore Company had failed. How- 
ever a corporate biographer of Smith and Moore, 
Hank Johnston, does not provide evidence to sup- 
port Curtis' recollection. 3 The company was in 
financial distress in 1892-94, however, but did not 
fail outright. What actually happened between 


Curits and Smith and Moore over his funding re- 
quest remains unanswered. 

In desperation Curtis turned to the Werner 
Company, holders of the exclusive concession 
rights, and sold "two or three thousand prints" 
to them. 5 By this time his ready cash had eb- 
bed to the point that he sought from a distant rel- 
ative living in Chicago (?) a loan sufficient only 
to get him back to California. 

Soon after his return, he read a newspaper art- 
icle concerning a great warehouse fire in Chicago. 
It was the Werner warehouse wherein all his prints 
(though not negatives) were stored. The company 
had no insurance. About that time someone offer- 
ed Curtis $3000 for the Noble negatives, but on the 
basis of the loose contract with Smith and Moore, 
Smith disallowed the sale. 37 

The demoralized 
Plains on the west 
tral Valley where 
land had been pure 
Colony failed near 
had sought to move 
never moved, Curti 
chased several 160 
the grain harvests 
had and his health 

Curtis moved to the Kettleman 
side of California's great Cen- 
he had 160 acres of land. The 
hased right after the Kaweah 
Sequoia National Park , when it 
elsewhere. Although the Colony 
s and some others privately pur- 
-acre tracts. He photographed 
but lost what little money he 
as well. 

He moved northward to San Jose a 
er ' s position in a spice mill there 
hid a great number of photographic 
some floor boards in the mill build 
after he moved to nearby San Franci 
city he worked at the J. A. Folger 
importers. He was credited with be 
four men who helped save the Folger 
ing during the great earthquake and 
Shortly after the holocaust he lear 
spice mill was to be razed. He was 
cover the negatives without inciden 

nd took a labor- 
in 1899. He 
negatives under 
ing. Soon there- 
sco. In that 
Company, coffee 
ing one of the 
mill from burn- 
fire of 1906. 
ned that the 
able to re- 
t, however. 39 

Curtis passed the remainder of his life in first 
Berkeley, and then later Pasadena, California. 
Little more has been published on his life than 
that related above. His position in the history 
of western Americana and in photography remains 
to be assessed. 

The camera he employed was a monstrosity of 45 
pounds and used glass negatives of 8 by 10 inch 
dimensions. To develop the negatives Curtis 
awaited nightfall and worked under a red lantern. 
The prints were made on albumin paper which had 
been sensitized by silver nitrate solution. 40 

The series of photographs accompanying this 
paper are, to the author's knowledge, all of the 
pictures still extant taken by Curtis of the Noble 
cutting. Each picture is described separately in 
the figure legends. Andrews has indicated that 
all pictures of the cutting operation are pub- 
lished in his work (1958). This is not the case. 
Curtis has said that he took 18 pictures of the 
operation. 4 ! Andrews shows much fewer. The pres- 
ent work is not complete either, but is believed 
to be as complete as is now possible. Curtis had 
indicated implicitly that he provided Harold Schutt 
of the Tulare County Historical Society with cop- 
ies of all remaining views. 42 

Other photographs accompany this narrative. 
Their origin is given in the figure legends and 
thanks for permission to reproduce them is given 
in the first footnote. 

At the Fair 
A German-American, Dr. T. W. Zaremba, made a 

FIGURE 1. The cutting crew and others at the base of the Noble Tree. The men are 1. to r. Sam Turk, Will 
Gwin, Dayton Dickey, Jesse Pattee, Capt. Jamison, Burr Mitchell, John Bodkins, Tom Gibson, Creed Archer. 
Identification after Historical Bulletin, 6:1950, Tulare County Historical Society. Schutt Collection. 



FIGURE 2. Another view of the gathered men and their supplies at the base of the Noble. Some can be 
identified from Figure 1, others are unknown. Schutt Collection. 

FIGURE 3. Capt. Jamison and Burr Mitchell c.t the base of the Noble. Schutt Collection. 



FIGURE 4. Initial scaffolding prior to first cut. Platform is at 25 feet above ground level. The men 
in the scene are not identified. It is possible the woman is Charles C. Curtis' wife. Schutt Collection. 

FIGURE 5. First cut at 52 feet and accomplished by double-bladed axe. It appears that the saw-cutting 
from opposite the axe is about ready to begin. Schutt Collection. 


FIGURE 6. The actual fall of the Noble. Although this reproduction is not of great quality, it is appar- 
ent that the original photo was very well done. To catch a tree in the act of falling without blur on 
the older, slow-speed films was the mark of a first-rate photographer. Schutt Collection. 

FIGURE 7. A^view of the upper portion of the Nokle as it lay on the ground (lower left) . The great split 
later to be "patched" can be seen on the left side of the tall stump (right of center) . Schutt Collection. 


FIGURE 8. Hollowing and sectioning the "upper room." 
winch line. Schutt Collection. 

Note careful handling of exhibit section on end of 

FIGURE 9. Time out for a bear hunt. Note slain American Black Bear, lower center, 
fied, but hunter (with rifle) may be Creed Archer. Schutt Collection. 


The men are not identi- 

FIGURE 10. The wheel section ("floor" of "upper room") has been removed. The platform is in the third, 
or lowest, position preparatory to cutting sections of the "lower room: of the exhibit. This picture, 
printed backwards, shows the split on the right side of the trunk above one man's head. Schutt Collection. 

FIGURE 11. Hollowing and sectioning the "lower room." Schutt Collection. 


FIGURE 12. A view of unloading marked sections to a waiting wagon. This scene was repeated many times 
throughout the cutting operation. Schutt Collection. 

FIGURE 13. The last wagon leaves with exhibit sections and the cutting is complete. Schutt Collection. 


FIGURE 14. Wagons with sections move down the specially constructed road on the way to the railhead. Re- 
printed from Andrews' Redwood Classic, with permission. 

FIGURE 15. The Noble exhibit as it appeared in the rotunda of the U.S. Government Building at the World's 
Columbian Exposition, 1893. After a German-language Fair guide publication of unknown origin. 


FIGURE 16. The U.S. Government Building. After Truman, History of the World's Fair, 

FIGURE 17. The Noble exhibit as it appeared while on the Mall in Washington, D.C. Note the cupola- 
bearing roof which at first protected visitors to its interior and later the stored lawn-care tools. The 
photo was taken by Harry Shannon, a newspaperman, in 1915. Note a portion of one of the original buildings 
of the Department of Agric ulture (built 1867) in the left background. Collection of the Columbia His- 
torical Society. 144 

FIGURE 18. A picture of Jesse Pattee as he looked in 1950 shortly before his death. Taken by Harold 
Schutt, the Noble Tree, known to modern visitors as the "Chicago Stump," is in the background. Schutt 


FIGURE 19. The Noble tree stump ("Chicago Stump"; in 1950 with Jesse Pattee in foreground. Note that the 
bark has almost completely sloughed off. A holocaust fire which swept the area in 1955 destroyed the lad- 
der which was never replaced. Schutt Collection. 


FIGURE 20. Charles C. Curtis. Reprinted from Andrews' Redwood Classic, with permission. 

FIGURE 21. The home of Charles C. Curtis and his family while recording the Noble cutting. The house is 
built upon a sequoia stump. Curtis is looking out the window with a daughter (Cecile?) by his side. Re- 
printed from Andrews' Redwood Classic, with permission. 


pronounced effort in June of 1884 to assemble a 
World's fair to commemorate the 400th anniversary 
of Columbus' discovery. The event was to be held in 
Mexico, but soon thereafter, Chicago was decided 
upon. The history of the Worlds' Columbian Expo- 
sition has been best treated in Truman and the 
interested reader is referred to this rather rare 
volume. 4 

The Visalia Daily Times began a 2-year long 
series on happenings at the "World's Fair" on July 
2, 1892. Great excitement then and throughout the 
reports the Times carried was in evidence, for 
Tulare County was to be well-represented. Not only 
would a great many exhibits be in the California 
Building, but the Noble section would have the dis- 
tinction of standing in the great rotunda of the 
United States Government Building, surely a great 

The agricultural prowess of Tulare County would 
be well-represented for 

it is only necessary that our 
[agricultural] advantages be 
known abroad to attract an enter- 
prising class of settlers and 
hasten development in which all 
are interested ." 

Finally on October 20, 1892, nearly a year later 
than originally planned, "an immense parade" of 
more than 50,000 participants opened the World's 
Columbian Exposition. 

The newspaper accounts in Tulare County indi- 
cated that considerable excitement had been genera- 
ted by the Fair. Local merchants capitalized on 
the event in a variety of ways. Merchants S. 
Sweet and Company in Visalia even offered hardbound 
books {The Columbian Worlds Fair Atlas, a guidebook) 
to their customers free of charge. 45 

Tulare County residents were exhorted to see the 
Fair and their county exhibit in person by purchas- 
ing a ticket on the Santa Fe Route: 

Ho For Chicago'. 
How to get to the World's Fair is the question 
The rai lroads want $100 
for the round trip ticket 
and there are many people 
in Tulare County who will 
have to stay at home. This 
will be too bad for everything 
worth seeing in the civilized 
world will be on exhibition 
in the windy city .... 

Therefore, one only had to contact the visalia 
Daily Times office to find out how, for three weeks' 
labor (doing what was not specified) , one could earn 
enough money for the trip or 

. . . to secure a $75 sewing 
machine , or a paid-up scholarship 
for six months in the Chestnutwood 
Business College at Santa Cruz 
[California] . 45 

So it was in Visalia and across the country, Go 
to Chicago. How many Visalia residents or vis- 
itors in general ever saw the Noble remains unknown. 

In addition 
raisins given 
ing the Fair, 
statuette. Ca 
likeness of Me 
been from wood 
such a sequoia 
of Grant Grove 
the authors co 
As well, the i 
mains mysterio 

to the 8,000 half-pound 
out by Fresno County exh 
they also displayed a my 
rved by "a young Italian 
rcury, the statue was sa 

taken from the Noble, 
-wood statuette exists i 
, Kings Canyon National 
llection) , but its origi 
dentity of the "young It 
us. The whole matter is 

boxes of 
i aitors dur- 
" in the 
id to have 
A picture of 
n the files 
Park (and in 
n is unknown, 
alian" re- 

in only one source and is not referred to else- 
where. 4 ^ 

The Fair passed into American history and the 
exhibits went back to their respective states in 
most instances. The Noble section, however, was 
destined for another, rather bizarre fate. 

Washington, D.C., and the Mysterious 
End of the Noble 

At the close of the Fair in 1893. the Noble was 
moved (by whom, by whose orders, or when exactly, 
are unknown) to Washington, D.C. There, on a con- 
crete slab poured for the purpose, stood the Noble 
Tree section from 1893 until 1932. With a cupola 
roof (see Fig. 17) it hosted visitors for many 
years. But, sometime prior to 1932, the spiral 
staircase was condemned and the whole exhibit 
closed to public use. 

For several more years the "structure" held 
lawn-care tools used by federal employees in their 
lawn-work near the old Department of Agriculture 
Building (erected in 1867) . The Noble stood near 
the northwest corner of Independence Avenue and 
Twelfth Street, southwest, on a mall proper. 
Various guidebooks of the day note it. 4 ° 

"The Rambler" 
October 31, 1915, 
Star said, 

The old r 
which has 
grounds s 
women of 
girls, ha 
Few perso 
wi tho ut s 
at 1 east 
glance , a 
one of th 
graphed o 

Several years 
in the artgravure 

(columnist Harry Shannon) in the 
issue of the Washington Sunday 

edwood tree trunk 

stood in the [Agriculture] 
ince mature men and 
Washington were boys and 
s taken on a mantle of ivy. 
ns pass this old relic 
topping to examine it and 
to pay it the tribute of a 
nd this long dead tree is 
e most frequently photo- 
bjects in the Capital. 

later it was a central picture 
Of the Washington Post. 

In 1932, it was dismantled and placed in storage 
in that condition in a steel shed at the United 
States Department of Agriculture Experiment Farm 
in Arlington, Virginia. It is not known why. 

In 1952, a manager of the Pacific 
pany of San Francisco wrote the Natio 
vice. He was inquiring about the di 
the section which he surmised was of 
1854. The NPS seemed uninformed in 
Noble and directed the letter to Dav 
Architect of the Capitol. He seemed 
formed but believed the Noble was mo 
ville, Maryland, when the Experiment 
was transferred to the Army for war 
never been substantiated. 

Lumber Com- 
nal Park Ser- 
sposition of 

a tree cut in 
1952 about the 
id Lynn, then 

as poorly in- 
ved to Belts- 
Farm land 
use. This has 

With all the revived interest generated by this 
author in determining the final disposition of the 
Noble section, no more information than was avail- 
able in 1956 has been forthcoming. In that year 
W. D. McClellan, originally of Tulare County and 
employed at United States Department of Agricul- 
ture in Washington, stated 

J have talked with several men at 
Beltsville who were acquainted with 
the Sequoia tree . . . Jimmy Taylor, 
our greenhouse foreman , Joe Rhodes and 
Johnny Alexander , our two fieldmen , 
were all familiar with it. They said 
they used it for tool storage .... 
After removal from [the Mall] to 
Arlington Farm it was enclosed in sheet 
metal . None of these men know of any 


order for moving it when [the Farm] 
was taken over by the Armed Forces for 
[construction of] the Pentagon , Rhodes 
and Alexander were the last two men 
to move material from Arlington Farm 
to Bel tsvil le . 

a ma 
di sp 
wen t 

us t b 
n at 
osi ti 

] cal 

wi th 

a gon 
re ar 

O n 
n t ia 

y chance 
Bel tsvill 
y been to 
on of the 
Colonel [ 
led someo 
asking w 
the tree 
to forget 
and to us 
a t he did 
may be bu 
to be une 
cheologi s 
the findi 
[sic ] at 

I did 
e who 
Id of 


in ch 

ne a t 

hat t 


he h 
e his 

I do 
ri ed 
ts wh 
ng of 








is C 

ad a 


n' t 


d by 

o wi 

a S 


k wi th 


t seems 

of the 
sked the 

7 u . / ■ I < • - 

know but 
r the 
11 spec- 
eq uoia 
tion'. 51 

It seems an inglorious end for such a noble 
tree. The record of misuse of these magnificent 
trees is astounding, however. Yet the Noble served 
an important educational purpose in its American 
odyssey, even though it would probably be alive 
today had there been no World's Columbian Exposi- 


1. For their many efforts the author wishes to 
thank Harold Schutt, Tulare County Historical 
Society; Robert Truax, Columbia Historical 
Society; Doris Randall and Marge Hamlin, 
Tulare County Library; the staff of Sequoia 
and Kings Canyon National Parks; authors Hank 
Johnston, Ralph W. Andrews, Kramer Adams; 
Elbert Little, U.S. D. A.-F.S. ; Helena M. Weiss, 
D. H. Nicolson and Conrad V. Morton, National 
Museum of Natural History-Smithsonian Institu- 
tion; Lllewelyn Williams, U.S . D. A. -Agricul- 
tural Research Service; Paul Spivey, U.S.D. A.- 
F.S.; David Mendengall, Chicago Tribune; 
Sylvia Arrigoni, Chicago Public Library; 
Kenneth B. Pomeroy, American Forestry Associa- 
tion; Larry L. Meyer, westways magazine; Re- 
cords History, U.S. Army; Lee S. Motteler, 
Rand-McNally and Company. 

2. Biology Department, Franklin College, Frank- 
lin, Indiana 46131. 

3. The binomial used here is that currently deemed 
most correct by R. J. Hartesveldt et al., in 
The Giant Sequoia of the Sierra Nevada, 
Washington, D.C., Government Printing Office, 
1975, page 24. 

4. Article, Visalia Daily Times, June- 7, 1892, 
p. 2. 

5. Article, visalia Daily Times, June 11, 1892, 
p. 1. 

6. Article, Visalia Daily Times, July 25, 1892, 
p. 2. 

7. Fern Gray, And the Giants Were Named (Three 
Rivers, California: Sequoia National History 
Association, n.d.), p. 12. 

8. Article, visalia Daily Times, date undetermin- 
able, but ca. August, 1892. 

9. Historical Bulletin (Visalia, California: 
Tulare County Historical Society, October, 
1950) , No. 6. 

10. Article, Visalia Daily Times, August 13, 1892, 
p. 1. 

11. Ma j . Ben C. Truman, History of the World's 
Fair (Chicago: Monarch Book Co., 1893). 

12. Dumas Malone, ed. , The Dictionary of American 
Biography (New York: Scribners, 1934) , p. 
539ff . 

13. Gray, And the Giants Were Named. 

14. The date of Aug. 12 has never been published 
elsewhere but is established by an article in 











I . 













the Visalia Daily Times, Aug. 13, 1892, p.l. 
Bulletin, p. 1. 
Ibid., p. 1-6. 

Hank Johnston, They Felled the Redwoods (Los 
Angeles: Trans-Anglo Books, 1966). p. 24. 
Gray, and Bulletin, p. 1. 

Article, Visalia Daily Times, Aug. 13, 1892, 
p. 1. 

Gray, and Fresno Weekly Republican , Aug. 19, 
1892, p. 8. However, others set the height 
at 50 feet. See Bulletin , p. 1. 

The U.S. Army has no records of a Captain Ja 
son and no other sources discuss him. His 
identity has remained a mystery. All others 
mentioned above in the figures were local 
lumberjacks or, in Mitchell's case, a local 
landowner-lumberjack . 

Harold Schutt, transcript of a wire-recorder 
interview with Jesse Pattee, Feb. 17, 1947. 
Xerographic copy of transcript in author's 

For example, in Ralph W. Andrews, Redwood 
Classic (New York: Bonanza, 1958), p. 109. 
Schutt, transcript. 

The transcript of an interview with C. C. 
Curtis done by J. R. Challacombe, no date, a 
popular expositor on sequoia lore, was pro- 
vided the author by Harold Schutt. Xerograph 
ic copy in the author's files. 
Challacombe, transcript. 

In Redwood Classic, Andrews states (p. 109) 
that Curtis came to California prior to 1880. 
This is apparently derived from an interview 
with one of Curtis' s daughters conducted by 
Andrews. This seems in error as Curtis him- 
self has stated, at least twice, that the yea 
was 1881. The corroboration is in the inter- 
view with Curtis by Challacombe and in a 
letter from Curtis to Schutt in 1950. Neithe 
of these sources was available to Andrews. 
The history of the colony has been treated in 
several publications, but see Douglas Hillman 
Strong Trees or Timber (Three Rivers, Cali- 
fornia: Sequoia Natural History Association, 
n.d.), p. 22ff. 
Challacombe, transcript. 

Johnston, They Felled, p 
Challacombe, transcript. 
Ibid . 

Andrews Classic, p. 109, 

Andrews, Classic, p. 109. 

Schutt, 1950 letter, and Challacombe, trans- 

Schutt, 1950 letter. 
Truman, History. 

Article, Visalia Daily Times, July 16, 1892, 
p. 1. 

Advertisement and copperplate, Visalia Daily 
Times, Jan. 12, etc., 1893, p. 3. 
Advertisement and copperplate, visalia Daily 
Times, Aug. 2, etc., 1893, p 3. 
Report of the World's Fair Commi ssion (San 
Francisco: State of California, 1893) , p. 43 
For example, a variety of Rand-McNally Guide- 
books of the Capitol mention it, viz., "A 
tower in the garden, composed of slabs with 
their foot-thick bark from one of the giant 
trees (Sequoia) of California, should not be 
neglected, for it represents the exact size 
of the huge tree, General Noble, from which 
the pieces were cut" (1900 edition; same word- 
ing for next 15 yearly editions) . Also, it 
is shown in map form in what There is to See 
in the United States Department of Agricultun 
(Washington, D. C. : U.S.D. A., ca. 1922), p. ( 

2 r >. 

and Challacombe 

and map (unnumbered pg.), a guidebook. Columbia Historical Society, Washington, D.C. 

49. Article and photograph, Washington Post, Mar. 51. W. D. McClellan, in a letter from him quoted 
13, 1921, artgravure section. in Los Tulares (successor to the Historical 

50. Edric C. Brown, letter to Director NPS, Jan. Bulletin) , Quarterly Bulletin of the Tulare 
11, 1952; and David Lynn, letter to Edrec C. County Historical Society, No. 26, Mar. 1, 
Brown, Nov. 6, 1952. Xerographic copies in 1956, p. 2. 

author's files from originals in files of 



Richard H. Hevly 

Naturally formed lakes are relatively rare 
phenomena among the wind and water sculptured pla- 
teaus, mesas, and canyons of the lower elevation 
and relatively more arid portions of southern 
Utah and adjoining northern Arizona. Not infre- 
quently, however, one encounters geological evi- 
dence of the former existence of lakes in such 
areas. The lakes have been formed in a variety 
of ways, but often they have been the result of 
damming of streams by lava flows and landslides. 
The sediments accumulated behind such dams often 
include fossils indicative of environmental con- 
ditions at the time of sedimentation. While ac- 
cumulation of lacustrine sediments in such basins 
varies with climatically and edaphically con- 
trolled runoff, sedimentation could potentially 
occur during intervals characterized by non-sedi- 
mentation or even erosion in alluvial environments. 
Hence, analysis of the fossil content of such 
lacustrine sediments affords not only the oppor- 
tunity to reconstruct the environment of deposition 
but also obtain paleoecological data on temporal 
intervals often lacking in nearby alluvial records. 

Recently an opportunity became available to 
examine sediments of five lakes in Zion National 
Park, Sentinal Slide Lake, Paria Lake (= Taylor 
Creek Slide Lake) , Beatty Lake (= Taylor Creek 
Bog), Trail Canyon Lake, and Coalpits Lake (Fig.l). 
These localities may be characterized as occurring 
between 4200 and 6300 feet elevation in Pinyon- 
Juniper Woodland with bordering pine-oak-fir for- 
est in adjoining coves and Chaparral on adjoining 
exposed south-facing slopes. Near streams which 
border the sample localities abundant riparian 
deciduous trees occur, including such taxa as wil- 
low (Salix), Cottonwood (Populus), ash (Fraxinus), 
alder (Alnus) , hackberry (Celtis) , walnut 
(Juglans) , birch (Betula) , and maple (Acer) . Fur- 
ther details concerning location and geology are 
given in Hamilton (1977). 

All of these lakes except Beatty Lake have 
breached their dams and erosion has exposed the 
lacustrine sediments in the banks of small washes. 
The sediments have been deposited in successive 
layers of coarse to fine grained texture with in- 
tervening layers of darker colored more organic 
sediments. Fossils sought included microfossil 
invertebrates, algal relicts and the spores and 
pollen of various plants as well as macrofossils 
such as invertebrates (e.g. snails and various 
arthropods), bones of vertebrates, and various 
plant fragments such as seed. Macrofossils were 
recovered by screening of the sediment and micro- 
fossils were recovered by chemical digestion of 
sediment samples . The sediment samples were 
usually obtained in a manner typical for arroyos : 
e.g. removal of 100 g samples from the principal 
strata in the cleaned face of a profile. At 
Beatty Lake, an extant landslide lake, a 24 
inch livingston piston corer was used to obtain a 
1.35m core subsequently divided into 5cm intervals. 

Samples were extracted by routine HCL-HF 

procedures for disaggregation and solution of 
carbonatesand silicates followed by acetolysis and 
KOH treatment for solution of soluble organics . 
A drop of acid-resistant residue from each sample 
was mounted on individual slides and covered by a 
#1, 22 x 40 mm coverslip. These preparations were 
scanned with a Leitz research microscope at a mag- 
nification of 450x. Macro- and microfossils were 
identified using standard references. Pollen pro- 
portions were determined and graphed. Selected 
proportions were graphed as departures from 
standardized means reducing the significance of 
variation in local vegetation for interpretation. 

Macrofossils (except mollusks in some localities) 
were scarce and poorly perserved as noted for land- 
slide lakes in Northern Arizona by Lance (m.s.) . 
Vertebrate bones (Bison, Thomomus , fish and bird) 
were recovered from Pleistocene Lake Trail Canyon 
(Hamilton, in press) . Tracks of Camel and a vari- 
ety of invertebrates including horsehair worm, ant, 
beetle, and aquatic insect larva as well as im- 
pressions of aquatic plants and a mussel were 
found in sediments of Coalpits Lake (Hamilton in 
press) . Mollusks were found in all lakes and car- 
bonized roots and stems of plants were encountered 
in some horizons particularly in the Holocene land- 
slide lakes. From a total of 204 mollusks obtained 
from the Holocene landslide lakes (Table 1), only 
three shells of aquatic snails (Gyraulus and 
Physa) and one valve of a clam (Sphaerium) were 
recovered. The Pleistocene lava lakes have thus 
far yielded 98 mollusks of which 87 are aquatic 
species (Table 1) . 

Microfossil remains were more abundant and more 
consistently recovered than previously studied 
landslide lakes, including pollen of seed plants, 
spore of Bryophytes (mosses and liverworts) and 
Pteridophytes (ferns and fern allies, e.g. 
Selaginella , spike moss, and Equisetum , horsetail 
or scouring rush) . The sediments proved to be 
largely devoid of microscopic aquatic invertebrates 
and algal remains but a few Diatoms were found and 
they were limnophilous periphytic species which 
thrive in water of high pH and conductivity sug- 
gesting shallow environments of high evaporatic 
potential (Table 2) . Only the pollen of seed 
plants were recovered in sufficient quantity from 
the relatively more fine textured and organic 
strata to determine the relative abundance of major 
types (presence only of other microfossils is in- 
dicated below; Fig. 2). Where possible, a pollen 
count was made of the first 200 grains encountered 
in contiguous non-overlapping rows. Percentages of 
the pollen counted in each sample were determined 
and are shown graphically below (Fig. 2), arranged 
according to provenience and stratigraphic rela- 

Department of Biological Sciences, Northern 
Arizona University. 

The pollen record can basi 
two major categories: arbore 
pollen and non-arboreal (NAP) 
grass pollen, The arboreal p 
types as pine (Pinus) , junipe 
(Ouercus), PAP (Picea, Abies, 
RAP (riparian tree pollen) . 
ther categorized as to size, 
ing mostly referable to pinyo 
larger grains are mostly refe 

cally be divided into 
al (AP) or tree 

or shrub, herb and 
ollen includes such 
r (Juniperus) , oak 

Pseudotsuga) and 
Pine pollen was fur- 
the smaller grains be- 
n pine, while the 
rable to yellow and 


white pines. Non-arboreal pollen included nu- 
merous pollen types such as Cheno-Ams (Chenopo- 
diaceae , gossfoot or lamb's quarter family + Amaran- 
thus or pigweed), Compositae (sunflower family, 
including Artemisia or sagebrush) , Gramineae (grass 
family) . Pollen of various aquatic herbs such as 
the Cyperaceae (Sedge family) , Umbeliferae (Carrot 
family) , Cruciferae (Mustard family) , Potamogeton 
(knotweed) , Typha (cattail) , Lemna (duckweed) , 
and Myriophyllum (water milfoil), was counted. 
The pollen sum included only terrestrial plants 
excluding the pollen of aquatic plants whose pro- 
portions were based on the total pollen count 
(Fig. 2) . 

Beatty Lake and Paria Lake: These lakes con- 
stitute a series of slide lakes on the same trib- 
utary in the northwest corner of Zion National 
Park. Beatty Lake (pollen series A) is an ex- 
tant pond situated in a narrow canyon vegetated 
by pinyon- juniper-oak woodland (ponderosa pine and 
fir in coves) at an elevation of 6300 feet. This 
pond is very shallow, weed filled, and bordered 
by riparian trees. The sediments were found to 
consist of more or less inorganic red clay to a 
depth of 130 cm. Pollen was obtained only from 
the relatively more organic strata at depths of 
0, 10-15, 20-25, 35-40, and 100-105 cm. Except 
for occasional fragments of the leaves and stems 
of aquatic herbs, no other macrofossils were 
found. The pollen record of the upper 40 cm. is 
relatively homogenous and may provide legitimate 
modern comparative samples for fossil samples 
from Taylor Creek. For that reason it will be 
described in detail below. 

The modern ephemeral weed-choked pond yields a 
pollen record with only 2 to 3.5% aquatic herb 
pollen. Although it is bordered by riparian trees, 
these pollen types constitute only 3.5 to 7.5% of 
the pollen record. The total arboreal pollen, in- 
cluding the riparian tree pollen, ranges from 46 
to 53% of the pollen recovered from terrestrial 
plants. The most abundant arboreal pollen types 
are pine and juniper, averaging 26 and 15% re- 
spectively (P/S ratio = 60% pine) and reflecting 
more or less the local vegetation type. Two- 
thirds or more of the pine is composed of small 
pollen grains probably referable to pinyon pine, 
thus also reflecting the local occurrence of that 
species. Although locally abundant near the lake, 
Oak {Quercus) pollen constitutes only 3.0 to 7.5% 
of the pollen form terrestrial plants, a pattern 
typical of western oaks. Nearby cove forests con- 
tain ponderosa pine, Douglas Fir (Pseudotsuga) , 
and White Fir {Abies). Large pine pollen grains, 
probably largely referable to ponderosa pine, con- 
stitute about one-third of the pine pollen ob- 
served, and pollen from other conifers of the cove 
forests contribute 1.5 to 3.0% of the total pol- 
len recovered. The remaining 36 to 54% of the 
pollen record is non-arbortal pollen composed pre- 
dominately of Cheno-Ams, Compositae, and 
Gramineae, averaging 11.5, 17.0 and 9.0% re- 

Tree-ring studies of a very old ponderosa pine 
which was partially buried by the rock-fall dam 
creating Beatty Lake indicate that this damming 
probably occurred about A.D. 1469 (Laing and 
Stockton, 1976). Approximately 25 meters of al- 
luvium have accumulated behind this dam suggest- 
ing an average sedimentation rate of 5 cm per 
year, closely agreeing with modern rates of 4.5 
cm per year, based on observations of sediment 
transport by Hamilton (in press). Therefore, the 
pollen record probably does not represent more 
than 30 years of record. The deepest sediments 
of Beatty Lake have not as yet been sampled, but 
those which have been sampled are sandier than the 

upper ones and contain 
ized by arboreal polle 
56 to 60%. Pine polle 
per less abundant than 
sediments. Pollen of 
herbs is more rare and 
more abundant than in 
at least one undated, 
corded in the sediment 

a pollen record character- 
n percentages ranging from 
n is more abundant and juni- 

in the upper 40 cm of pond 
riparian trees and aquatic 

the pollen of Compositae 
the upper sediments . Thus , 
drier, cooler episode is re- 
s of Beatty Lake. 

The entire 13-meter section of Paria Lake 
(area = .04 mile 2 ; Hamilton, in press; pollen 
series C and B) differs from the stratigraphically 
superimposed Beatty Lake in the predominance of 
sandy to gravelly sediments containing a variety 
of fossils, including organic lenses of herbaceous 
plant hash, stumps and limbs of trees, gastropods, 
and a number of different kinds of microfossils 
such as pollen of seed plants, bryophyte and 
pteridophyte spores, and algal relicts (e.g. 
diatoms) . The sediments of this lake are dated 
by Cm dates of 2880 ± 200 and 3610 + 300 B.P. 
near the top and bottom of this section respec- 
tively (Eardly 1966; Hamilton, in press). 

The upper sediments of the Paria Lake are 
particularly characterized by the occurrence of 
numerous lenses of plant hash approximately 2.5 
to 7.5 cm. in thickness alternating with much 
thicker lenses of sand. This zone proved to be 
particularly productive of pollen unlike the ma- 
jority of slide lake sediments sampled here and 
elsewhere in the Southwest (Lance, n.d.). The 
pollen record of the section, which was obtained 
almost exclusively from the organic seams, ex- 
hibited variation of composition and proportion 
of arboreal pollen both less and greater than that 
observed in the upper 40 cm of Beatty Lake. The 
surface sediments of Paria Lake are very sandy 
and gravelly and extremely eroded, but one sample 
(C-l) from very near the top of the section has 
yielded enough pollen for a 100-grain count. This 
sample is remarkable because of the great reduc- 
tion of AP (8%) and increased relative abundance 
of NAP (89%) , particularly of Compositae and 
Poaceae (55 and 21% respectively) . The inferred 
environment based on modern AP pollen proportions 
from northern Arizona would suggest a warmer, 
drier environment than presently found in the st 
area; however, one that provided adequate moistu 
for the persistence of both aquatic herbs and 
riparian trees (3%) . Although the AP record is 
very limited, the slightly greater relative abun- 
dance of juniper than pine may possibly indicate 
an open woodland or savanna with local over- 
representation of Compositae and Gramineae in the 
pine behind the lower dam. 

The uppermost sample from Paria Lake differs 
strikingly from the next lower sample (B-l) 
which has high proportions of arboreal pollen 
(53.5%), particularly of juniper and PAP. Pol- 
len of the Gramineae, Cyperaceae and Typhaceae 
are also more abundant than in the upper levels 
of Beatty Lake. Samples B-2 through B-8 are all 
remarkable because they contain less arboreal 
pollen than in the upper levels of Beatty Lake, 
while samples B-13, B-14 and B-17 contain more 
arboreal pollen than the surface or near-surface 
samples of the pond. 

The implications of these AP proportions 
when compared to studies of modern pollen re- 
covered from various plant communities of northern 
Arizona is that the environment surrounding Paria 
Lake has undergone change from woodland to savan- 
na and back again to woodland on several occasions, 
Such apparent changes of AP proportions could, 
however, be strongly influenced by NAP production. 



Studies of the modern pollen rain in northern 
Arizona have shown that analysis of AP composi- 
tion (Pine-Juniper ratios and Pine-size ratios) 
can provide a technique of reconstructing environ- 
mental change independent of the influence of 
over-representation of local NAP (Hevly, 1968). 
When examined in this way (Fig. 3) only the 100- 
105 cm sample from Beatty Lake exhibits more pine 
in its pine-juniper ratio than found in the wood- 
land vegetation of the Colorado Plateau; however, 
most samples from Paria Lake exhibit less pine and 
more juniper than found in the modern woodland 
pine-juniper ratios of the study area or on the 
Colorado Plateau. Samples with high AP and high 
pine proportions resemble near pine canopy situa- 
tions in woodland, while the samples with high 
AP and high juniper resemble near juniper canopy 
situations in woodland. Samples characterized by 
low AP and with low pine and high juniper propor- 
tions probably reflect Juniper-Savanna (samples 
B-2 through B-8) . Thus, the analysis of arboreal 
pollen generally supports the inferences developed 
on the AP/NAP ratios alone. 

Pollen, spore, algal relicts (including sever- 
al diatom types) and occasional shells of aquatic 
snails definitely indicate that at least the or- 
ganic sediments of Paria Lake represent lacustrine 
depositional environments. At no time does this 
pond appear to have been deep, since plants 
tolerating only shallow water are present through- 
out the record and intervening sandier lenses con- 
tain only shells of land snails. Regionally at 
least two periods more xeric than at present. 
Plants very much like those occurring locally ap- 
pear to have been present throughout the several 
thousand years spanned by this very fragmentary 

Sentinal Slide Lake: Sentinal Slide Lake 
(Area = .772 mile ,• Grater, 1945; Hamilton, in 
press) is located in the south fork of the Virgin 
River at an elevation of about 4,300 feet and is 
surrounded by pinyon- juniper-oak woodland and 
chaparral with a riparian woodland the adjoining 
upland from which water and sediments are denied 
contain ponderosa pine, white and Douglas Fir and 
Aspen bordering the permanent stream. Sediments 
of Sentinal Slide Lake like those of Paria Lake 
contain abundant sands; however, in the Emerald 
Pool (--D — series) and to a more limited extent 
in the Birch Creek (--E — series) exposures of this 
lake, greenish or grayish clays and silts are com- 
mon and often contain organic residues. At the 
Birch Creek locality a red silt near the top of the 
section (E-4) has been C 14 dated at 3600 ± 400 
(W3371 Hamilton, in press). 

In Sentinal Slide Lake only the organic clays 
and silts contain pollen, suggesting that pollen 
deposition and preservation occurred only in the 
depositional environments with low sediment trans- 
port capability. Pollen samples from Sentinal 
Slide Lake generally exhibit increasing relative 
abundance of pine and PAP with depth and increas- 
ing proportion of large pine pollen relative to 
small. Pollen of Juniperus , Quercus , RAP, 
Ephedra, Cheno-Ams , and Compositae generally di- 
minish with increasing depth. The only departure 
from this general trend thus far detected occurs 
in sample E-4 where pine values drop to j%, some 
14% less than in modern samples. Compositae and 
and Gramineae pollen increase in relative abun- 
dance 27 and 6% respectively over modern values. 
Pollen of Cyperaceae, Typha, Potamogeton , Lemna, 
Myriophyllum, and Equi setum , Selaginel la , ferns, 
bryophytes , and algal relicts also increase during 
this interval. The sediments of this horizon 
are brown silts including carbon films of aquatic 

plants while underlying and overlying this are 
thick strata of sand. 

Pollen, spore, 
clam shells defin 
textured sediment 
sent lacustrine d 
ning about 3600 + 
ment as reflected 
became increasing 
ward in time to a 
(1977) estimates 
Paria Lake. 

and algal relict and aquatic 
itely indicate that the finer 
s of Sentinal Slide Lake repre- 
epositional environments. Begin- 
400 B.P., the regional environ- 
by the terrestrial pollen types 
ly more mesic progressing back- 
t least 4000 B.P. which Hamilton 
to be the initiation date of 

The fluctuating presence and absence of pollen, 
spores, and algal relicts along with the varia- 
tion of sediment texture suggest considerable in- 
stability in Paria and Sentinal Slide Lakes. 
That is, there appear to have been pulses of depo- 
sition characterized by fine-textured sediments, 
fossils, and apparently sufficient persistence of 
moisture that aquatic plants become established 
and accumulated more rapidly than they decayed. 
The periods characterized by coarser textured 
sediments and absence of fossils may represent 
relative brief periods of rapid deposition per- 
haps under conditions of flooding or at least 
conditions which were generally unsuitable for 
preservation of organic materials. Flushing of 
land snails form nearby surfaces probably occurred 
under such conditions. These lakes seem rarely to 
have had appreciable depth as pollen of deep wa- 
ter tolerant species was recovered abundantly in 
only a few strata of Sentinel Lake; however the 
alternating lacustrine and non-lacustrine deposi- 
tional record may parallel in part the changing 
water level detected in other Post-Pleistocene 
Southwestern lakes (Hevly, 1964; 1974). These 
lakes provide a Post-Pleistocene record often miss- 
ing in alluvial sections because of erosion. At 
least twice during the Post-Pleistocene, environ- 
ment conditions as indicated by the above pollen 
data have undergone significant fluctuation sug- 
gestive of the Post-Pleistocene environmental 
changes widely documented in western North America 
(Adam, 1975; Hevly 1974; Madsen , 1972; Mehringer 
and Warren, 1976; Phillips and Van Devender , 1974; 
Sercelj and Adam, 1975.) 

Trail Canyon and Coalpits Lak 
were formed by lava damming duri 
and the two samples described he 
cellent comparative sample for P 
palynology and paleoecology . Th 
these lakes is different from th 
Slide Lake on the nearby North F 
River, since their watersheds at 
no fir and little Ponderosa Pine 
obtained recently contain about 

(over half of which is composed 
grains) , 16% juniper, 10% oak, 8 
Compositae, and 13% Gramineae. 
samples contain about 11% PAP, 7 

(over half of which is composed 
grains) and 8% grass pollen. Th 
ference revealed by this compari 
percentages clearly indicates th 
environmental change that had oc 
elevations in Zion National Park 
treat of the last continental ic 
pollen record of these lakes ind 
magnitude and direction comparab 
in nearby areas of Nevada and No 
as indicated by both macroscopic 
data (Hevly and Karlstrom, 1974; 
1961; Mehringer, 1967; Van Deven 
1971; Spaulding, 1977; Wells and 

es : These lakes 

ng the Pleistocene 

re provide an ex- 


e provenience of 

at of Sentinal 

ork of the Virgin 

present contain 
Modern samples 
1% PAP, 20% pine 
of small sized 
% Cheno-Ams, 20% 
The Pleistocene 
6% pine pollen 
of large sized 
e tremendous dif- 
son of pollen 
e magnitude of 
curred at lower 

since the re- 
e sheets. The 
icates changes of 
le to these found 
rthern Arizona 

and microscopic 

Martin et . al . , 
der and King, 

Berger , 1967) . 


TABLE 1. Fossil mollusks recovered by Hamilton (in press) from Holocene and 
Pleistocene Lakes, Zion National Park. Ecological data from Bequaert and Miller 
(1973) . 



Southwestern Trail Canyon Fossil Occurrence 
Ecology & Coalpits Paria L. Sentinel L. 

Cochlicopa lubrica ( Mul ler ) 

Discus cronkhitei (Newcomb) 

feet in drift 
in seasonally 
most canycs 

feet in humid 
to moist condi- 
tions under 
rock or wood 
or in litter 

Discus shimekii (Pilsbry) 7,200-12,100 

Euc onulus fulvus (Muller) 5,000-10,500 


Microphysula ingersollii (Bland) 7,000-11,000 
Oreohelix strigosa (Pilsbry) 3,000-8,000 

II ii II ii II II 

Succinea avara (Say) 3,500-7,200 

-■— ■ ii ii H n u n 

Vallonia perspectiva (Sterki) 3,500-8,700 




6 + 1 



Aquatic Clams 

Pisidium sp. 

Sphaerium ( Musculium ) partumeium 

Species not 
known but 
genus common 
in Southwest 
ranging from 
3,500 to 9,500 
feet in streams, 
and lakes and 
small ponds. 

Species not 
listed for 
Other species 
range in ele- 
vation from 
3,000 to 5,000 
in creeks and 


Aquatic Snails 

Ecol o gy 

Trail canyon 
& Coalpits 

P ^ria T,. .Sen tina 1 L . 

Physa sp. 

Species not 
known but 
genus common in 
Southwest, rang- 
ing from 1,000 
to 8,700 feet in 
small flowing 
streams and 
stagnant water. 

Lymnae (Stagnicola) bul imp ides 

He! i soma tenue (Dunker) 
Planorbella tenuis 

Gyraulus parvus (Say] 

Subspecies not 
listed for South- 
west. Other sub 
sp. range in ele- 
vation from 4,850 
to 6,900 feet in 
springs, small 
pools, and other 
moist habitats. 

Ponds & lakes us- 
ually below 6,000 
feet but ranging 
up to 8,700. 

Springs, ponds 
& lakes between 
3,000 and 
8,250 feet 

1 + 4 

17 + 54 

TABLE 2. Diatoms from Paria and Beatty Lakes. Identifications by David Czarnecki, 
Dept. Bio. Sci. N. Ariz. Univ., Flagstaff. Ecological inferences based on Lowe (1974) 


Ecological Inferences 

Fossil Occurrence 
Beat ty I.. Paria L. 

Cv^bella cistula (Hemp.) Grun. 

E pithemia turgida (Ehr.) Kutz 

Fragilaria vaucheriae (Kuts.)Pet. 

Navicula sp. 

Alkaliphilous (opt. pH 8) 
Limnophilous to 

1 imnobiontic 
Peri phytic 
Cosmo pol itan 

Alkaliphilous (opt. pH 8) 
Saproxenous to beta 
mesosaprobic ' 
Peri phytic 
Common Southwest sp. 

Alkaliphilous (opt. pH 

6.5-9.0) grows best in 

waters of high 

Limnophilous to 

Peri phytic 
Cofnnon Southwest sp. 

Too poorly preserved for 
identification Species of 
this genus usually inhabit 
standing water but a few 
occur in flowing water as 
well. Many species are 
periphytic becoming conrnon 
in the summer and fal 1 . 


Core A 

+(Loc. B) 




FIGURE 1. Map of Zion National Park showing the 
locations of five Holocene or Pleistocene Lakes: 
Beatty Lake (=Taylor Creek Bog) , Paria Lake (= 
Taylor Creek Slide Lake) , Sentinal Slide Lake, 
Trail Canyon Lake and Coalpits Lake. 





Taylor Creek Bog Surface 
TaylorCreek Bog 125 cm( 
TaylorCreek Bog 25cm 
TaylorCreek Bog 40cm | 
TaylorCk Bog100-105cm I 
Tavlor Creek c 
TaylorCk Slide Lake B 1 | 
TaylorCk Slide LakeB2 
TaylorCk Slide LakeB 4 | 
TaylorCk /Slide LakeB 6 | 
TaylorCk /Slide LakeB ? | 
TaylorCk Slide Lake B 8 
TaylorCk SlideLakeBg | 
TaylorCk SlideLakeBg I 
TaylorCk /SlideLakeBg 
TaylorCk 'Slide LakeB 14 
TaylorCk SlideLakeB 17 | 

Arboreal Pollen 

10 20 30 40 50 60 70 

■ ■ I 
I I 





Emerald Pool/ Trail-modern I 
Sentinal Slide/Lake 0-00 I 
Sentinal Slide/Lake X 
SentinalSlide/Lake E-4 | 
Sentinal Slide/Lake E-2 | 
Sentinal Slide / Lake D- 7 
Trail Canyon Lava Lake 3 
Coal Pits Wash/Lava Lake 

Non Arboreal Pollen 

o£ ™ 

) MM 





Aquotic Herbs 


1 — ' r 

FIGURE 2. Pollen diagram of Beatty Lake (=Taylor Creek Bog) , Paria Lake (=Taylor Creek slide 
Lake) , Sentinal Slide Lake, Trail Canyon lava Lake and Coalpits lava Lake. The near equivalency 
of AP/NAP and AP composition ratios of Holocene Sentinal Lake sample D-7 and those from the 
Pleistocene Lakes seemingly suggests profound environmental change during the Neoglacial in 
Zion National Park exceeding that observed elsewhere in the Southwest. Alternatively the data may 
reflect redeposition of arboreal pollen from the Kolob Plateau with increased runoff water during 
particularly mesic interval since Martin (1963) has observed pine pollen to be transported in 
froth of s'liraner flood water in southern Arizona. Such transport is not a universal phenomenon, 
however, since pine pollen was not recovered in summer flood froth in northern Arizona (Hevly, 1970) 
perhaps as a result of in flood fragmentation or oxidation prior to, during or after flooding or 
even removal by earlier floods. Therefore the data may reflect both expanded pine forest and an 
unknown amount of exaggeration due to redeposition since the records exceed other Southwestern 






AP Departures 


L.Pine | Pine 




from Modern 

I P/j Ratiol 

Composition Ratiol 


o « 

-20 +20 

-20 o +20 

-20 +20 


Bealty Lake 

A. D. 1975- 

| = Taylor Creek Bogl 
















+ .2 





S 5 


i t 

— 3 


in r 


B.P. 2000- 






+ o 




fl, (9 


5 > 



w — 



























■o > 



— 3 


y> - 







> 10,000- 

Trail Canyon Lake 




• o 

4) <Q 

>10n 0- 

Coalpits Wash L. 




> i 

— 3 

FIGURE 3. Summar 
The palynological 
in northern Arizo 
pollen proportion 
cant at the 5% le 
Studies of growth 
that breaching of 
deep has been cut 
about 6.3 cm per 
sediment. More 1 
youngest sediment 

y diagram of major palynological data rela 

data are plotted as departures from mean 
na and southern Utah, the stippled bars re 
s from the mean values (percentages which 
vel) . Both Paria and Sentinal Lakes leave 
in a tree whose roots were exposed by ero 
that lake occurred by A.D. 1740 (Laing an 
in 230 years indicating an erosion rate o 
year. Carbon-14 dates indicate that this 
ikely the duration of time to fill is seve 
s to accumulate have been eroded away. 

ting to moisture 
pollen proportio 
presenting two s 
exceed the stipp 

breached their 
sion of sediment 
d Stockan, 1976) 
f these unconsol 
lake took at lea 
ral hundred year 

and geolog 
ns of moder 
tandard dev 
led areas a 
dams and ha 
s of Paria 

A channe 
idated sedi 
st 730 year 
s greater s 

ical trends, 
n woodlands 
iations of 
re signifi- 
ve been eroded. 
Lake indicate 
1 130 meters 
ments of 
s to fill with 
ince the very 


The Holocene lakes of Zion National Park were 
formed by landslides during relatively more cool- 
moist intervals. Sedimentation in the resulting 
basins was primarily attributable to accumulation 
of fine to coarse grained sediments washed into 
the basin during periods of run-off from summer 
storms and melting snow. Differential sedimenta- 
tion of the particles has resulted in successive 
layers of sand, silt and clay. The basins con- 
tinued to accumulate sediments during intermittent 
periods of relatively more warm-dry intervals and 
alternating cool-moist intervals until the sedi- 
ments filled the basins and the dams were breached. 
The duration of these lakes has been estimated to 
be generally less than a millenium. 

The Holocene lakes appear never to have had 
appreciable depth as indicated by the scarcity of 
pollen of aquatic plants tolerant of deep water 
and the abundant pollen of aquatic plants tolerant 
of shallow water. As indicated by the ecological 
requirements of the scarce fossil diatoms, the 
water of the lakes appears to have been alkaline 
(pH 8-9) and of high conductivity further suggest- 
ing shallow environments of high evaporative 
potential. These lakes were also generally ephem- 
eral as indicated by the scarcity of fossil snails 
requiring permanent moisture and the abundance of 
terrestial snails. Nevertheless there were some 
intervals during which standing water existed in 
these basins since diatoms, aquatic snails, and 
pollen of aquatic plants accumulated in darker 
colored, more organic seams among the alternating 
sands, silt, and clay layers. 


Analysis of the sediments from these Holocene 
lakes has yielded for the first time a record of 
conditions during non-depositional intervals recog- 
nized in alluvial sections elsewhere in the South- 
west. Environmental conditions changed signifi- 
cantly during the deposition of sediment in these 
basins, erosion, or non-deposition outside and 
sometimes inside the basins apparently correspond- 
ing to intervals characterized palynologically 
by diminished arboreal pollen while periods of 
alluviation and landslides were characterized by 
increased arboreal pollen. By comparison to 
modern pollen samples obtained from a variety of 
modern plant communities diminishing arboreal pro- 
portions should reflect increasing aridity while 
the opposite would be true of increasing arboreal 
proportions (Fig. 3). The interpretation so 
derived is in agreement with trends noted else- 
where in the Southwest. 

The Pleistocene lakes are strikingly different 
from the Holocene lakes. The Pleistocene lakes 
formed behind lava dams rather than landslides. 
The snails are more abundant and many more of them 
are species requiring permanent moisture, suggest- 
ing greater permanence to these Pleistocene lakes 
than the Holocene lakes. Thicker accumulation of 
sediments also suggest greater duration for these 
lakes prior to breaching of the dams . 


Appreciation for 
this study is expres 
tory Association and 
for released time fo 
travel . The author 
field and laboratory 
W. Hamilton who both 
to the better unders 
herein. Partial res 
sented at the First 
tional Parks (A.I.B. 
1976 and the Arizona 
Vegas, Nev. , April, 

financial support to undertake 
sed to the Zion National His- 

to Northern Arizona University 
r analysis, interpretation and 
has greatly benefited from the 

assistance of J. Ward and 

contributed through discussion 
tanding of the data presented 
ults of this study were pre- 
Conference on Research in Na- 
S.) - New Orleans, LA, Nov.. 

Academy of Science - Las 


ADAM, D. P. 1975. A late Holocene pollen record 
from Pearson's Pond, Weeks Creek landslide, 
San Francisco Peninsula, California. J. Res. 
U. S. Geol. Survey. 2 (6) : 721-731 . 

BEQUAERT, J. C. and W. B. MILLER. 1973. The 
Mollusks of the Arid Southwest. Univ. Ariz. 
Tucson. 271 p. 

EARDLY, A. J. 1966. Rates of denudation in the 
high plateaus of southwestern Utah. Geo. Soc. 
Amer. Bull. 77:777-780. 

GRATER, R. K. 1945. Landslide in Zion Canyon, 
Zion National Park, Utah J. Geo. 50(20) :116- 

HAMILTON, W. L. in press. Holocene and Pleisto- 
cene Lakes in Zion National Park. Proceedings 
of First Conference on Research in National 
Parks. (A.I.B.S.) 

HEVLY, R. H. 1964. Paleoecology of Laguna 
Salada. in: P. S. Martin et al . , Chapters 
in the prehistory of Eastern Arizona, II. 
Fieldiana (Anth.) 55:1-261. 

. 1968. Modern pollen rain in North- 
ern Arizona. J. Ariz. Acad. Sci. 5 (2) : 116-127 . 

. 1970. Botanical Studies of a 

Strange Jar Cached at Grand Falls Arizona 

1974. Recent Paleoenvironments and 
Geological History at Montezuma Well. J. 
Ariz. Acad. Sco. 9(2):66-75. 

HEVLY, R. H. and T. N. V. JARLSTROM. 1974. 

Southsest Paleoclimate and Continental Cor- 
relations, in: Geology of Northern Arizona. 
(T.N.V. Karlstrom and others, editors) . 
Geological Soc. Amer. (Rky. Mt. Sect.) Field 
Guide 2:257-295. Flagstaff. 

LAING, D. and C. W. STOCKTON. 1976. Riparian 
dendrochronology; a method for determining 
flood histones of ungaged watersheds. final 
report OWRT Project # A-058-An3. by laboratory 
of tree-ring Research, University of Arizona. 

LANCE, J. F. n.d. Lake and Moqui Canyons. Un- 
published manuscript. 

LOWE, R. L. 1974. Environmental Requirements 

and Pollution Tolerances of Freshwater diatoms. 
EPA - 670/4-74-005. Cincinnati, 334 pp. 

MADSEN, D. B. 1972. Paleoecological investiga- 
tions in Meadow Valley Wash, Nevada. In: D. 
D. Fowler (ed.), Great Basin Cultural Ecology: 
A symposium Desert Research Institute Publi- 
cations in Social Sciences. 8:57-65. 

MARTIN, P. S. 1963. The Last 10,000 Years. 
Univ. of Ariz Press Tucson 87 pp. 

1961. Rampart cave coprolite and ecology of 
the Shasta Ground Sloth. Am. J. Sci. 259:102- 

MEHRINGER, P. J. JR. 1967. Pollen analysis of 
the Tule Springs Site, Nevada. in: Pleisto- 
cene Studies in Southern Nevada. H. M. 
Wormington and D. Ellis (eds.) Nevada State 
Museum Anth. Pap. No. 13:129-200 

Marsh, dune and Archaeological Chronologies, 
Ash Meadows, Amargosa Desert, Nevada. Nevada 
Archeo. Survey Research Report 6. 

PHILLIPS, A. M. Ill and T. R. VAN DEVENDER. 1974. 
Pleistocene packrat middens from the Lower 
Grand Canyon of Arizona. J. Ariz. Acad. Sci. 
9(3) : 112-119. 

SERCELJ, A. and D. P. ADAM. 1975. A late Holo- 
cene pollen diagram from near Lake Tahoe, 
El Derado County, California. J. Res. U.S.G.S. 
3(6) :737-745. 

SPAULDING, G. 1977. Late Quaternary vegetational 
change in the Sheep Range, Southern Nevada. 
J. Ariz. Acad, Sci. 6 (4 ): 240-244 . 

VAN DEVENDER, and J. E. KING. 1971. Late Pleis- 
tocene vegetational records in western Arizona. 
J. Ariz. Acad. Sci. 6 (4) : 240-244 . 

WELLS, P. V. and R. BERGER. 1967. Late Pleisto- 
cene History of Coniferous Woodland in the 
Mohave Desert. Sciences. 155:1640-1647. 



Phillip L. DeBord' 


Thousands of people visit Yellowstone each year 
to see its geysers, travertine terraces and asso- 
ciated mineral pools, animal life, plant life, and 
the beauty of its terrain. Many backpack to ex- 
perience Yellowstone's beauty in a deeper way. 

However, few people ever see one of the most 
unique sights in all of the world-the "Fossil For- 
ests" in Yellowstone. 

Holmes' (1878) diagram (Figure 1) illustrates 
quite well the suggested sequence of events that 
produced the petrified forests. A forest was 
growing in a broad valley. A volcano erupted. 
The heavy ash covered the leaves and limbs of the 
trees, choking out life and causing them to fall 
off the main trunk. Breccia from the volcano 
caused the local streams to overflow their banks, 
inundating the forest. The slow-moving subariel 
or subaqueous breccias delivered the death blow to 
the forest, surrounding the trunks of the standing 
trees and preserving them in their position of 
growth. After approximately 200 years conditions 
were conducive for reforestation. A new forest 
grows to maturity and the sequence is repeated. 
Therefore, what is preserved for us is a series 
of forests (as many as 50) one on top of the 

A number of studies of modern pollen (Lichti- 
Federovich and Ritchi 1968, Tinsley and Smith 
1974, Webb III 1974) illustrate that there is a 
relationship between the pollen spectra and the 
vegetation of an area. Webb III (1974) states, 
"Maps of the relative abundances of pollen data, 
therefore, provide direct information about maps 
of the relative abundances of vegetational data." 

The purpose of this study is to apply Webb's 
hypothesis to the unique setting of the Gallatin 
"Fossil Forests" in an effort to reconstruct the 
paleoclimatology , paleoecology , and depositional 
history of these forests. 


The study area is located approximately three 
miles from Specimen Creek bridge (which crosses 
U.S. Highway 191) on the southernmost flank of 
Big Horn Peak. 

The stratigraphy of the area as described here 
is the result of the work of Smedes and Prostka 
(1972). A 335' vertical section with twenty-two 
forests has been described. The section starts 
at the contact between the Fortress Mountain 

Member and the Daly Creek Member of the Sepulcher 
Formation and descends to the tree line and 
associated ground cover. The section was started 
at the contact because of the visual distinct- 
ness between the two members. The Fortress 
Mountain Member is a light bluish gray color 
while the Daly Creek Member is a dark reddish 
brown to gray. Thus, any subsequent investigator 
can readily find the beginning of the section. 
The age of the section is late early Eocene to 
early middle Eocene, i.e., late Wasatchian to 
early Bridgerian provincial stage. 

Petrologically, the Daly Creek Member in 
this section is composed of volcanic andesitic 
conglomerate in a pumice and ash matrix of 
alluvial origin (Smedes and Prostka 1972) . 
Interbedded with the conglomerate are stratified 
lenses of tuffs containing organic material. 


Field Techniques 

The section was described by the following 
procedure: 1) using a Brunton Compass and start- 
ing at the contact between the two members of the 
formation, I descended to the first organic zone; 
2) this organic zone contained a vertical trunk 
with its base in the organic zone and therefore 
was designated unit 1 or forest 1; 3) the vertical 
depth of the unit was recorded; 4) the organic 
zone was traced across the slope and described 
in relation to thickness, organic content and any 
"unusual" characteristics, i.e., disappearance 
and reappearance, evidence of water deposition, 
and splitting into two or more zones; 5) vertical 
trunks were typed with a hand lens as being pine, 
sequoia, or deciduous, if field identification 
was not possible, wood samples were taken and 
identified from thin sections in the laboratory; 

6) it was noted whether vertical trunks possessed 
roots and if so the condition of the roots; 

7) the diameter of all vertical trunks, horizonal 
logs and diagonal logs was recorded as well as the 
intervening distance; 8) palynology samples were 
taken from the organic zone; 9) the distance 
between samples and the type tree near the collec- 
tion site was recorded and 10) the grading, 
bedding, and sorting of the unit was recorded. 

The lateral extent of each forest was from the 
approximate area of the fault on the west end 
of the section around the corner on the east end 
to the tree line unless the breccia promontories 
were too precipitous (lower part of the section). 

This study is concerned with units 1, 5, 6, 9, 
and 10 because they contain an adequate number of 
trees and discrete forest types. 

This work was done at the Department of 
Biology, Loma Linda University, and was supported 
by a grant from Geoscience Research Institute, 
Berrien Springs, Michigan. I wish to thank Glen 
F. Cole, Supervisory Research Biologist, for his 
cooperation, Marilyn Emtage for making the thin 
sections and keying out the wood that could not be 
field-identified, Bill Hughes for his help with the 
photography, and Dr. A. J. Roddy and Kathy Ching 
for their help with the manuscript. 


Department of Biology, Loma Linda University, 

Loma Linda, California 92354. 

Laboratory Techniques 

Because of the qu 
thirty-one of the ni 
collected could be u 
Only six of the thir 
ing preservation qua 
be utilized by assig 
major category. Thr 
or spores, were coun 
the entity into one 
1) clubmoss or fern, 
type, 4) angiosperm, 

ality of preservation only 
nety palynology samples 
sed for making pollen counts, 
ty-one samples had outstand- 
lity; however, the rest could 
ning the pollen or spore to a 
ee hundred grains, pollen 
ted from each sample placing 
of the following categories: 
2) bisaccate, 3) taxodiaceous 
and 5) indeterminable. 


FIGURE 1. W. H. Holmes' diagram of a section of the Amethyst Mountain "Fossil Forest. 
Reproduced from Holmes (1878) . 

TABLE 1. Tree composition (vertical and horizontal). 

Unit 1 Unit 5 Unit 6 Unit 9 Unit 10 Total 
















2 \ 

2 6 







I lftd»t«rmln»bU Dli-actlon: W« a i to Bwl UNIT t 

FIGURE 2. Composite histogram of unit 1. 





iZ.l SR . 2 

FIGURE 3. Composite histogram of unit 5. 




FIGURE 4. Composite histogram of unit 6. 

> Mr«tU». Watt to I* 

FIGURE 5. Composite histogram of unit 9. 






SttnpU Otr*cclon: W..i to E«* 

FIGURE 6. Composite histogram of unit 10. 

TABLE 2. Percent trees vs. percent pollen. 

Unit 1 

Unit 5 

Unit 6 

Unit 9 

Unit 10 

Sequoia trees 






Taxodeaceous pollen 

4 r , 

3 2 

4 5 



Pine trees 


9 5 


Bisaccate pollen 






Deciduous trees 

4 3 

6 9 



Angiosperm pollen 







All specimens illustrated in this study were 
photographed with the Polaroid attachment of a 
Zeiss Ultraphot II binocular research microscope. 
All photomicrographs were taken on Type 55 
positive/negative Polaroid film. These photo- 
graphs were taken with an oil immersion (100X) 
planachromat objective using Nomarski inter- 
ference contrast to emphasize the surface orna- 
mentation. Enlargements were made on Kodak 
Polycontrast rapid RC paper and prints were 
developed in Kodak Dektol developer. All photo- 
graphs other than palynomorphs were copied from 
35mm Ilford FP4 film and developed as above. 

All figured specimens are shown on the plates 
at a standard magnification of 1000X to illustrate 
the relative size differences. 


At this time there are over 130 forms in. the 
palynoflora. As with other Eocene palynofloras 
(Fisk and DeBord 1974, Leopold 1974) it indicates 
a flora with great ecological and taxonomic 
diversity (Plates 1-3). Taxonomic study of the 
palynoflora is still in progress. 

A composite histogram of each unit was 
constructed from the relative abundance of each 
category in each sample. Therefore, any category 
can be traced across the unit. 

Unit 1 or forest 1 is a mixed forest contain- 
ing twenty-three trees (Table 1) . Looking at 
the angiosperm category (Figure 2) and tracing it 
from the west end of the unit to the east end it 
appears to be quite consistent from sample to 

sample except for samples 1H-3 (unit 1, sample H, 
slide # 3) and U-5. These two samples are from 
a split area in the organic zone. 1H-3 is from 
the lower zone and 1J-5 is from the middle zone 
just above 1H-3. There is a dramatic change in 
the angiosperm component from a mixture to a 
dominance of Alnus in the lower zone and uiwus in 
the middle zone. At the present time I have no 
explanation for this phenomenon. Overall the 
other categories in this unit appear to be quite 
consistent from sample to sample. It should be 
noted that no statistical tests have been applied 
to any of the preliminary data presented in this 

Unit 5 (Figure 3) is a forest consisting of 
only sequoia and deciduous trees and no pines 
(Table 1). Each category is fairly consistent 
across the unit with several categories showing 
greater concentrations between samples, but over- 
all the spectra is quite consistent. The pres- 
ence of pine pollen (bisaccate category) in this 
unit could possibly be explained by wind trans- 
port from forests growing at a higher elevations. 

Unit 6 (Figure 4) is another forest with only 
sequoia and deciduous trees (Table 1) . Although 
preservation in this unit was extremely poor, 
there were three samples that could be utilized, 
which gave a spread across the unit. Although 
sample 6E-3 does have an abundance of taxodeaceous 
type pollen, all categories are consistent 
across the unit. 

The palynoflora of unit 9 (Figure 5) indicates 
a forest dominated by Alnus. But this is a pine 
forest (Table 1) . Because of the prolific nature 
of pine the pollen profile should be extremely 
high in this forest, i.e., 60-70% of the 


palynoflora (McAndrews and Wright, Jr. 1969). 
Only one sample shows a percentage that indicates 
deposition beyond the forest edge. 

Unit 10 (Figure 6) is another mixed forest 
(Table 1) . Of all of the units this one is the 
most consistent from sample to sample. It is 
interesting to note that this unit having seven 
more pine trees than unit 9 has considerably more 
pine type pollen. 

Table 2 shows the comparison between the 
relative abundance of tree type and pollen type. 
Units 1, 5,6 and 10 show a close correlation. 
Units 1 and 10 show an underrepresentation of 
pine type pollen. Unit 9 is quite anomalous, 
overrepresented in deciduous type pollen, 
especially in Alnus , and extremely under- 
represented in pine type pollen. 


The preliminary data from units 1, 5, 6, and 
10 are in general agreement with the results 
from studies of modern forests. However, unit 
9 cannot be explained by any modern analog and 
represents a striking anomaly. Further study 
may resolve this problem. 


FISK, L. H. and P. L. DEBORD . 1974. Plant micro- 
fossils from the Yellowstone fossil forests: 
preliminary report. Geol . Soc . Am. Absts. with 
Programs 6(5):441-442 (Abst.). 

HOLMES, W. H. 1878. Report on the geology of 
the Yellowstone National Park. U.S. Geologi- 
cal Survey of the Territories, Annual Report 
no. 12, pt. 2, pp. 1-57. 

LEOPOLD, E. G. 1974. Pollen and spores of the 

Kisinger Lakes fossil leaf locality. pp. 49-66. 
in: An early middle Eocene flora from the 
Yellowstone-Absaroka volcanic province, north- 
western Wind River Basin, Wyoming. H. D. Mac- 
Ginitie. University of California Publications 
in Geological Sciences 108. 103 pp. 


Recent pollen assemblages from western interior 
of Canada. Rev. Palaeobot. Palyn. 7:297-344. 

MCANDREWS, J. H. and H. E. WRIGHT, JR. 1969. 
Modern pollen rain across the Wyoming basins 
and the Northern Great Plains (U.S.A.). Rev. 
Paleobot. Palyn. 9:17-43. 

SMEDES, H. W. and H. J. PROSTKA. 1972. Strati- 
graphic framework of the Absaroka volcanic 
supergroup in the Yellowstone National Park 
region. U. S. Geol. Surv. Prof. Paper 729-C. 
33 pp. 

TINSLEY, H. M. and R. T. SMITH. 1974. Surface 
pollen studies across a woodland/heath transi- 
tion and their application to the interpreta- 
tion of pollen diagrams. New Phytol. 73:547- 

PLATE 1. Clubmosses and ferns. (All figures 1000X.) 
1-Cardio angulina 2-Selaginel la 3-Del toidospora 
4-undulatisporites 5-Unknown trilete 6-Unknown tri- 
lete 7-?Cicatricosisporites S-Verrucatosporites 
9- Laevigatospor ites ovatus 10-?Schi zaea . 


PLATE 2. Gymnosperms. (All figures 1000X.) 1-Cyca- 
iopites 2-Unknown monosulcate 2-Taxodium hiatipites 
1-Taxodiaceae-pollenites hiatus 5-Sequoia lapilli- 
oites 6-7Thuja or Junipcrus. 


PLATE 3. Angiosperms. (All figures 1000X.) 1-Fre- 
montia 2 and 3-Tilia 4-Momipites coryloides 5-Nyssa 
6-Carya 1-Vlmus-Zelkova S-Planera 9-Alnus 10-Acer 
11-Salix 12-Liliacadites 13-Unknown 14-Unknown 15-? 
Populus 16-Uex 17-Unknown 18-Rhus 19-Unknown. 



Ronald H. Hofstetter and Frances Parsons 


This study was undertaken to assess the current 
status of sawgrass: its growth characteristics, 
vigor and distribution, and the environmental fac- 
tors affecting it, in the Florida Everglades. 


The Everglad 
land in the con 
about 10,000km 2 
southern Florid 
head 1971) . Hi 
southern end of 
mangrove fringe 
basin, it is bo 
Coastal Ridge a 
Rise and the Bi 

es is the largest contiguous peat- 
tinental United States, occupying 

or roughly one-third the area of 
a (Davis 1943, Egler 1952, Craig- 
storically it extended from the 

Lake Okeechobee southward to the 
Lying in an elongated shallow 
rdered on the east by the Atlantic 
nd on the west by the Immokalee 
g Cypress. 

Sawgrass, Cladium jamaicense Crantz, is the 
dominant plant of the Everglades, accounting for 
up to 70% of the vegetative cover (Loveless 1959) . 
Sawgrass occurs as a monoculture; or mixed with 
other gaminoids, herbs and aquatics, and it can 
be found in virtually every community in southern 
Florida . 

The Everglades has a "Tropical Savanna" climate, 
in the sense of Koppen (Hela 1952) , with 
alternating wet and dry seasons, a mean January 
temperature of 20 °C and a mean July temperature of 

The mean annual precipitation ranges from 110 
to 165cm, varying with location. Between 60 and 
80% of the annual precipitation occurs during the 
wet season from May to October (Thomas 1970) . 
Rainfall is associated mostly with convective 
activity, is highly variable in quantity and 
distribution, and is frequently accompanied by 
gustiness and lightning. Precipitation in the dry 
season occurs mostly as light steady rains 
associated with the passing of cold fronts, and 
lightning is rare. 

The Everglades basin is the lower part of the 
Kissimee River-Lake Okeechobee - Everglades drain- 
age system (Parker et al. 1955). The average slope 
of the Everglades is 2.8cm per kilometer (Hartwell 
1970) , so there is a general tendency for water to 
flow southward as a slow, broad sheet during the 
wet season when most of the area is flooded. 
Local heavy precipitation, however, may temporar- 
ily alter local flow patterns. This fresh-water 

We are grateful to the National Park 
for financial support from 1971 to 1975 i 
of contract EVER-N-48 as part of its cont 
to the South Florida Environmental Projec 
tude is also expressed to USDA-Forest Ser 
Research Agreement to continue ihe work o 
grass, particularly the spot sampling in 
Numerous people have contributed, but spe 
thanks are due to Dr. T. R. Alexander, Uni 
of Miami, and Mr. L. Bancroft, Management 
gist, Everglades National Park. 
Department of Biology, University of 

Coral Gables, Florida 33124. 

n the form 
t. Grati- 
vice for a 
n saw- 

Miami , 

head and seaward flow prevents intrusion of salt 
and brackish water up the estuaries into the 
Everglades. During the dry season the water 
table drops and surface water appears only in the 
deepest sloughs and gator holes. At the peak of 
the dry season, some surface sediments may be- 
come dusty dry. The decreased fresh-water head 
and flow permit brackish water to move inland of 
the mangrove fringe. 

The Everglades is underlain by limestone of 
Pleistocene and Miocene age (White 1970) . The 
sediment of part of the Everglades is muck and 
peat, derived mostly from sawgrass. It is deepest 
at the north end and shallower southward. This 
Histosol is classified as Typic Medihemist and 
Medisaprist (McCollum et al. 1976) . In the 
southernmost part of the Everglades, muck exists 
only in the deepest sloughs and most of the sur- 
face sediment is calcareous marl, deposited 
through the action of algae and bacteria. Marl 
forms an impervious seal over the limestone bed- 
rock . 

Fire has been a natural force affecting all of 
the ecological communities of southern Florida, 
promoting the graminoid communities and pine- 
lands, preventing invasion by mangroves and hard- 
woods (Robertson 1953) . As most of the natural 
fires are of lightning origin, they are common in 
the wet season, when sediment ana plant roots are 
protected by moisture, and uncommon for most of 
the dry season (ibid) . 

Man has interfered extensively with the natural 
processes of hydrology and fire in the Everglades 
to the extent that destructive, man-caused, dry- 
season fires occur almost annually in this impor- 
tant natural area. 


Sawgrass is a coarse sedge consisting of a 
firm stock, from which roots and rhizomes, leaves 
and inflorescences grow. The apical meristem 
lies at the top of the stock, continuing to pro- 
duce the leaves, which are "V"-shaped in cross- 
section and are serrated along the edges and the 
underside midrib (Yates 1974) . Vegetative 
growth continues year-round, but slows somewhat 
during the dry season (Fisher 1971, Forthman 
1973) . In flowering the primary meristem is 
carried upward at the apex of the developing 
inflorescence. Following fruiting in July- 
August, the culm dies, but tillers arising 
vegetatively from rhizomes continue to develop. 
Density of inflorescences varies greatly from one 
area to another and from one year to the next, 
(Alexander 1971) and controlling factors are un- 
known. Large numbers of seeds may be produced, 
but viability is usually low (ibid) and varies 
from year to year. Vegetative reproduction is 
common and seedlings are uncommon within 
established sawgrass communities (ibid) . 

The leaves have a high proportion of aerenchyma, 
which in mature green leaves and standing dead 
leaves may be involved with gas exchange between 
the submerged plant parts and the atmosphere. 


Conway (1938) demonstrated that Cladium mariscus , 
a related species, oxygenates its roots by diffusion 
through the aerenchyma of its leaves. Parsons 
(in prep.) has evidence that there is also some 
aeration of the rhizosphere of C. jamaicense. 

Where flooding is continuous sawgrass forms 
tussocks and the meristem may develop above the 
sediment surface (Alexander 1971, Yates 1974). 


There is considerable heterogeneity in species 
composition and in sawgrass vigor even within 
individual communities. This fact, along with the 
high cost of reaching remote areas by helicopter, 
required modification of methods commonly used in 
vegetation studies. 

To supplement the informa 
analyses of permanent transe 
reported in part by Forthman 
and Parsons (1975), one-time 
were taken in sawgrass commu 
throughout the Everglades (s 
tions) . Water depth, cover 
and density of inflorescence 
square meter quadrats were c 
ment surface. In the labora 
were separated, and the tota 
and the length and width of 
each sawgrass culm were dete 
and green leaves of sawgrass 
other species were separated 

tion gained from re- 
cts begun in 1971 and 

(1973) and Ilofstetter 
, meter-square samples 
nities at 26 sites 
ee Fig. 1 for loca- 
of plant components, 
s were recorded, and 
lipped at the sedi- 
tory plant components 
1 number of leaves, 
the longest leaf in 
rmined. The brown 

and individuals of 
, dried at 105°C, and 

The above-ground biomass at site 8 contains 
31,000kcal • m -2 ash-free, dry weight of living 
and standing dead sawgrass (Ilofstetter and 
Parsons 1975) . The ash content of this matter is 
between 3.0 and 7.4% of dry weight (ibid). 

The values of sawgrass characteristics recorded 
in Table 1 should not be used alone as an indica- 
tion of the overall quality of sawgrass in those 
areas sampled or in the entire Everglades . Many 
more samples and determinations of the percentage 
area occupied by the different communities and of 
the vigor of the vegetation within each community 
would be necessary . 

Insufficient time has passed to 
annual growth patterns for individu 
culms, but some possible trends are 
One culm had 6 leaves and an averag 
of 78cm in February, £.nd in October 
leaves with an average leaf length 
During that period this culm produc 
total leaf length. If the leaves a 
and 13mm wide, this represents a pr 
to 92g dry weight and 335 to 440 kc 
In that period 920cm (length) of gr 
died and 860cm (length) of standing 
including 7 entire leaves, became 1 

Decline of Sawgrass 

al sawgrass 

developing . 
e leaf length 

it had 10 
of 18 5cm. 
ed 2400cm of 
re between 10 
oduction of 64 
al. of energy. 
een leaves 

brown leaves, 


Our observations over the past 8 years indi- 
cate that sawgrass is continuing to decrease in 
distribution and in stature throughout the entire 
Everglades system. There are insufficient data 
on past conditions, however, to quantify these 
observations . 

Sawgrass Character is tics 

Values for the selected characteristics of saw- 
grass from the meter-square sampling sites are 
presented in Table 1. The maximum values from 
transects studies between September 1971 and 
October 1976 are: length of longest leaf: 317cm, 
site 15; width of longest leaf; 21mm, site 16; 
leaves per culm: 10.1, site 16; density: 67 
culms • m~2, site 23. Individual values for each 
transect at several times are recorded elsewhere 
(Forthman 1973, Hofstetter and Parsons 1975). 

The largest value for standing crop (at site 
8) is the result of high values of both green and 
standing dead material. This sample was selected 
from a small stand that had an unusually high 
density (55 culms • m~ 2 ) or robust culms (average 
values of 283cm and 20mm for length and width' 
respectively of longest leaves and 10.2 leaves 
per culm) . These values are much higher than 
those typical of sawgrass in that area. Several 
sites (5, 11, 14, 16) had similar values for above- 
ground green biomass. At site 5 the individual 
culms were small, but thei.v density was high, 
whereas at site 16 there was lower culm density, 
but the culms were robust with more numerous and 
more robust leaves. The culm density at site 14 
was more than double that at site 16, and the 
length and width of the longest leaves were com- 
parable, but there were fewer leaves per culm at 
site 14. This indicates, then, that measurement 
of no one or two characteristics of sawgrass will 
serve to predict standing crop of green matter. 

The plot of width against length of longest 
sawgrass leaf (Fig. 2), however, reveals a linear 
relationship with a correlation coefficient of 
0.942. Biomass of standing litter reflects both 
the rate of green matter production and of litter 
production, seasonality, and is greatly affected 
by time since burning. 

The causes of sawgrass change appear to be a 
combination of natural and man-caused phenomena, 
although these are not yet fully understood. The 
primary controlling factor apparently is 
hydroperiod (Robertson 1953, Loveless 1959): the 
depth, timing, and duration of flooding. The 
canals, levees, 3,500km2 of Water Conservation 
Areas, and water management practices influence 
all of the Everglades (Fig. 1) . In the southern 

(downstream) ends of the Conservation Areas the 
depth of surface water and/or period of flooding 
has increased. Here sawgrass is being replaced 
by cattail (Typha) and other hydrophytes 

(Loveless 1959, Alexander 1971, McPherson 1973, 
Alexander and Crook 1975) . In much of the 
historical Everglades the depth and period of 
flooding has decreased and the surface sediments 
are excessively dry and/or dry for unusually long 
periods. This problem has been intensified by 
several unusually dry years between 1970 and 
1975. In these drier areas sawgrass is being 
displaced by more mesophytic species, for 
example, Panicum hemitomum , Andropogon spp. and 
especially by woody plants, such as Myrica 
cerifera, and Baccharus spp. (Robertson 1953, 
Loveless 1959, McPherson 1973, Alexander and 
Crook 1975) . 

Sawgrass has also increased in stature and 
area in some locales. In the wetter parts of the 
Conservation Areas where sawgrass has been dis- 
placed by more hydric species, it has replaced 
more mesic species on somewhat higher ground. 
Where the hydroperiod has been shortened saw- 
grass has displaced more hydric species in the 
sloughs (McPherson 1973, Alexander and Crook 
1975). Overall, however, sawgrass has experienced 
a net loss (Robertson 1953, Loveless 1959, 
Hofstetter and Parsons 1975) . 

As a consequence to the reduced hydroperiod, 


V^ 5 


FIGURE 1. Map of southern FLorida showing major natural and anthropogenic features, and locations 
of meter-square sampling sites. 

TABLE 1. Average values (and standard deviation) of sawgrass characteristics at selected sites 
throughout the Everglades. 







Inf lor. 





Cover (%) 








- 2 






m 2 












14 (3) 

6.8 (1.6) 














13 (4) 

7.9 (2.6) 














17 (8) 

8.5 (3.3) 











4-11 I -76 


N. Trans. 



15 (6) 

6.1 (2.8) 





2 30 






Site 2-3 



13 (7) 

6.7 (3.8) 











Site 10 



12 (6) 

5.8 (2.7) 











NT* altlllere exc. 



17 (5) 

8.7 (2.5) 











14- IV-76 

b)tlllers Inc. 



14 (6) 

7.0 (2.6) 





27 50 




14- IV-76 



Site 1 



20 (6) 

10.2 (3.7) 
















16 (5) 

8.7 (2.0) 









7- IV-76 





14 (3) 

9.9 ( - ) 
















13 (4) 

10.6 (2.9) 














11 (4) 

8.0 (2.1) 











Br-2W a) Cladlua 



13 (7) 

8.3 (4.8) 






3 50 

3 7 





b) Typha 



13 (4) 

5.9 (2.2) 














Site 1 



16 (5) 

7.2 (2.0) 




10 50 





14- IV-76 






13 (4) 

8.4 (2.4) 









16- 11-76 





16 (6) 

10.1 (3.8) 







16- 11-76 


L67B Typtaa 



13 (3) 

6.8 (1.9) 

2 6 





8 50 










10 (3) 

6.4 (1.8) 














10 (4) 

6.8 (2.1) 




2 500 








15 (4) 

8.8 ( - ) 



2 550 



4 9 

9- 11-76 





11 (4) 

7.5 (. :.) 







10 50 









13 (5) 

8.4 (3.1) 













PHO-Trans. 1 



11 (3) 

6.3 (1.8) 











PHO-TriDS. 3 



12 (2) 

6.1 (2.1) 













PHO-Trans. 3 



10 (4) 

6.8 (2.4) 













Loatmao R. 



8 (3) 

7.5 (1.9) 







16- IV-76 

D.M. - dead Kiddle, I.e. ruin ehoae Interior Ihih arc brown while eoee of 
the exterior leivee are still Brean 

O.D. - ovan-drlad at 10S°C 

8»e ttit for further aiplaoatloDa 



WIDTH )4 . 





R = 0.94 

••• ^r • • 
*•••• • • • 

• ••••• • • • 

•• ••• 

• —~m— •• • •* 

• • • • • ••• 

• • «• • • • 

T" ""» ' t ■ t- -[ t — r -r 
200 250 

—J 1 1 1 1 1 — 

300 350 


FIGURE 2. Relationship of length to width of the longest sawgrass leaf per culm. 









k •' 



• — _' 

? _.-- 


•-s. 25. 

IOC'2 4 6 8 10 12' 2 4 6 8 10 12' 2 i 6 8 10 121 2 i 6 8 10 12 1 2 i 6 8 10 
71 7J 73 74 73 76 











— I — 



r ■•■ 

80 ^ 





MP* - 

■- ? -. 

-. 16 



■ « 

10 I2'2 4 6 8 10 12 2 < 6 8 10 12 2 4 6 8 10 lF2 4 6 8 10 I2 T 2 4 6 
71 72 73 74 75 76 


FIGURE 3. Changes in length of longest leaf, density, and cover of sawgrass with time 
at two sites in Everglades Natinonal Park. 


the extension and duration of brackish water has 
increased in the lower ends of the fresh-water 
wetlands, and sawgrass is being replaced by salt 
marsh species (Robertson 1953, Alexander and 
Crook 1975) . Another serious consequence of 
reduced hydroperiod is the increased frequency 
and severity of fires, especially ground fires 
(Robertson 1953, Loveless 1959, Hofstetter and 
Parsons 1975) . These are primarily incendiary 
(arson) , dry-season fires that kill sawgrass and 
other species that normally benefit from surface 
fires when they are protected by moisture during 
the wet season. 

Another problem has been the invasion and 
displacement of sawgrass and other native species 
by exotic plants, especially cajeput {Melaleuca 
quinguenervia) (Alexander and Crook 1975, Hofstetter 
and Parsons 1975) . This woody species is rapidly 
colonizing mesic sites and natural sawgrass 
communities (when the sediment is adequately ex- 
posed) . Melaleuca tolerates burning, has 
semiserotinous fruit, and is generally promoted 
by fires. Dense forests of this eucalypt are 
forming, mostly along the eastern edge of the 
Everglades. All of the Everglades system is 
threatened and will succumb eventually if preven- 
tive action is not taken soon. 

Besides being replaced by other species saw- 
grass has been dying in excess of normal, 
particularly in the interior of the densest 
stands. This condition, called decadent sawgrass 
(Hofstetter 1973) , was first observed in winter 
1970, by Alexander and Hofstetter and has been 
spreading rapidly throughout the Everglades. In 
some of the still-living culms the leaves are 
unusually short and chlorotic. This appears to 
be a nutritional problem that is not yet under- 
stood. More commonly seen, however, are normal 
culms with dead innermost leaves. In some cases 
death of the youngest leaves and meristem appears 
to result from fungal infection. In some instances 
normal elongation of the innermost leaves is 
hindered and the leaves impact accordian style 
for unknown reasons. Most commonly, however, 
death appears to result from the action of a 
boring larva, tentatively identified as 
Scirpophaga perstrialis (Hofstetter and Parsons 
1975) . At one study site 31% of the culms, and 
at another site 14 culms • m~2 (12% of culms 
present) were so killed. 

to the meristem is likely and especi 
the muck (peat) is dry enough to bur 
must be avoided when the water table 
submerging the fire-pruned culm will 
its death (Alexander 1971, Forthman 
ing of the developing inflorescence 
in loss of these seeds and death of 
H. Werner of Everglades National Par 
comm . ) recommends burning sawgrass i 
February with 10 to 15cm of surface 
long-term effect of annual burning o 
burning has not yet been determined. 

Hydroperiod Management 

ally when 
n. Burning 
is rising as 
result in 
1973) . Burn- 
will result 
the culm, 
k (Pers. 
n January and 
water. The 
r of time of 

Sawgrass appears to require an annual period of 
dry down (Loveless 1959) , but the return of the 
hydroperiod to a natural situation is unlikely 
and probably impossible for much of the Ever- 
glades system. Nevertheless some efforts are 
being made or (at least) are being planned to im- 
prove this situation. 


Length and width of the longest leaf per culm, 
numbers of leaves per culm, culm density and 
standing crop of sawgrass vary greatly through- 
out the Everglades and within individual commu- 
nities. There appears to be a linear relation- 
ship between the width and length of the longest 
leaves that holds for sawgrass over much of its 
range. Standing crop of sawgrass can not be 
accurately determined except by clipping and 
drying . 

Sawgrass appears to be declining in general 
distribution, dominance, and vigor within the 
Everglades. Some of this change can be attri- 
buted, at least in part, to altered hydroperiod, 
to excessively hot dry-season fires, to dis- 
placement by cajeput, to insect damage, and 
possibly to nutrient conditions. Studies on all 
of these factors are continuing. 

There is little reason to suspect that the 
vegetational changes in the Everglades will not 
continue, and consequences of a continued de- 
cline in sawgrass are not known because the role 
this plant plays in the Everglades is not fully 

Figure 3 shows 
in two of the stu 
Pa-Hay-Okee Trans 
from 1971 to 1972 
studied post-burn 
decadent conditio 
have been re-exam 
1975) . With deca 
may remain robust 
and width of leav 
cover. Sawgrass 
covering, with va 
density increasin 

the changes that have occurred 
dy sites (16 = L67-S, 25 = 
ect 3) since 1971. The values 
are from Forthman (1973) who 
growth responses. Since the 
n began to appear these transects 
ined (Hofstetter and Parsons 
dence, while some of the culms 
, most culms decline in length 
es, culm density, and sawgrass 
at L67 now appears to be re- 
lues of leaf length and culm 


Prescribed Burning 

Prescribed burning is now common practice in 
southern Florida, mostly in the early tc middle 
part of the dry season to reduce fuels for fire 
control (Hofstetter and Parsons 1975) . Before 
the decadent sawgrass condition becomes advanced, 
burning appears to delay its onset, but does not 
appear to prevent or to reverse the developing 
condition . 

Prescribed burning must be avoided when damage 


ALEXANDER, T. R. 1967. Effect of Hurricane 

Betsy on the southeastern Everglades. Quart. 

Jour. Fla. Acad. Sci. 30(l):10-24. 
ALEXANDER, T. R. 1971. Sawgrass biology related 

to the future of the Everglades ecosystem. 

Soil and Crop Sci. Soc. Florida Proc . , 31:72- 

ALEXANDER, T. R. , and CROOK, A. G. 1975. Recent 

and longterm vegetation changes and patterns 

in South Florida. Part II. Final Report. 

Mimeo. Report to Dept. of Interior, National 

Park Service. May 197 5. 
CONWAY, V. M. 1937. Studies in the autecology of 

Cladium mariscus R. Br. III. The aeration of 

the subterranean parts of the plant. The New 

Phytologist. 36:64-96. 
CRAIGHEAD, F. C. 1971. The trees of South 

Florida. Vol. 1. The natural environments 

and their succession. Univ. Miami Press, 

Coral Gables, Fl. 212 p. 
DAVIS, J. H. 1943. The natural features of 

southern Florida. Florida Geol . Surv. Bull. 

25. 311 p. 


DINEEN, J. W. 1972. Life in the tenacious Ever- 
glades. In Depth Report. 5(1). Central and 
Southern Florida Flood Control District. 

EGLER, F. E. 1952. Southeast saline everglades 
vegetation, Florida and its management. Vege- 
tatio 3(4-8) :213-265. 

FISHER, J. B. 1971. Inverted vascular bundles in 
the leaf of Cladium (Cyperaceae) . Bot. Jour. 
Linnean Soc . 64 (3) : 277-293 . 

FORTHMAN, C. A. 1973. The effect of prescribed 
burning on sawgrass, Cladium jamaicense Crantz, 
in South Florida. Masters Thesis. University 
of Miami, Coral Gables, Florida. 

HARTWELL, J. H. 1970. Some aspects of the avail- 
ability of water from the Everglades to the 
Everglades National Park, Florida. Open File 
Report 70007, U.S. Geol. Survey, Tallahassee, 
Fl. 36 p. 

HELA, I. 1952. Remarks on the climate of South 
Florida. Bull. Marine Sco. of Gulf of Mexico 
and Carribbean Sea. 2 (2) : 438-447 . 

HOFSTETTER, R. H. 1973. Effects of fire in the 
ecosystem - an ecological study of the effects 
of fire on the wet prairie, sawgrass glades, 
and pineland communities of South Florida. 
Report EVER-N-4 8 to U. S. Dept . Interior, 
National Park Service. NTIS Accession No. 

HOFSTETTER, R. H. and PARSONS, F. 1975. Effects 
of fire in the ecosystem. An ecological study 
of the effects of fire on the wet prairie, 
sawgrass glades, and pineland communities of 
South Florida. Part II. Final Report EVER-N- 
48, Department of Interior, National Park Ser- 
vice, May 1975. 

LOVELESS, C. M. 1959. A study of the vegetation 

of the Florida Everglades. Ecol. 40(1): 1-9. 
MCCOLLUM, S. H., CARLISLE, V. W. , and VOLK , B. G. 

1976. Historical anc current classification of 
organic soils in the Florida Everglades. Soil 

and Crop Sci. Soc. Florida Proc . 35:173-177. 
MCPHERSON, B. F. Vegetation in relation to 

water depth in Conservation Area 3, Florida. 

U. S. Geol. Survey, Open-file Report 73025, 

Tallahassee, Fl. 
PARKER, C. G., FERGUSON, G. E., LOVE, S. K., et 

al . 1955. Water resources of South Eastern 

Florida with special reference to geology and 

ground water of the Miami Area. U. S. Geol. 

Surv. Water Supply Paper 1255. 
ROBERTSON, W. B. 1953. A survey of the effects 

of fire in Everglades National Park. U. S. 

Dept. Interior, National Park Service. Mimeo 

Report. Homestead, Fl. 169 p. 
THOMAS, T. M. 1970. A detailed analysis of 

climatological and hydrological records of 

South Florida with reference to man's influence 

upon ecosystem evolution. Tech. Report 7 0-2. 

Rosenstiel School of Marine and Atmospheric 

Science, Univ. Miami, Miami, Fl . 
WHITE, W. A. 1970. The geomorphology of the 

Florida peninsula. Florida Bur. Geol., Geol. 

Bull. No. 51. 
YATES, S. A. 1974. An autecological study of 

sawgrass, Cladium jamaicense , in southern 

Florida. Masters Thesis, Univ. of Miami, 

Coral Gables, Fl . 



Frances Parsons" 


Phosphorus and nitrogen are essential macro- 
nutrients in plants. Phosphorus is believed to be 
often a limiting factor in freshwater ecosystems. 
The chemical forms in which nitrogen exists and 
is taken up by plants growing in waterlogged 
organic sediments is not completely understood. 
As sawgrass, Cladium jamaicense Crantz, occurs in 
virtual monoculture in many large areas of the 
Everglades of southern Florida, it is important 
to study possible influences on sawgrass of changes 
in these vital nutrients caused by seasonal 
climatic changes. 

Phosphorus has been shown to have a short 
residence time and a rapid turnover rate in some 
freshwater bodies (Hutchinson and Bowen 1950, 
Rigler 1956, Pomeroy 19 60) . Pomeroy (1960) 
pointed out that the turnover rate of phosphate is 
probably more important to community well-being in 
freshwater ecosystems than in the concentration of 
phosphorus . 

In natural plant communities, usable forms of 
nitrogen result from microbial transformations of 
organic nitrogen compounds of plant and animal 
origin and from nitrogen fixation. Usable forms 
of nitrogen to plants are nitrate, an anion, and 
ammonium, a cation. Nitrite, an intermediate 
state in the geochemical cycle of nitrogen, is of 
interest because it is toxic to plants under 
certain conditions (Bingham et al. 1954, Court 
et al. 1962). Climatic changes through the 
seasons affect the size, composition, and 
activity of the microbial populations responsible 
for these changes (Viets 1965, Davy and Taylor 
1974) . Climatic changes also affect the nature 
and quantity of the organic material upon which 
the microorganisms live. The annually alternating 
submergence and exposure of the sediments that 
occur over much of the Everglades would be ex- 
pected to have an effect on the chemical forms 
that nitrogen takes in the sediment. When Ever- 
glades sediments are drained for agricultural 
crops, nitrification proceeds rapidly (Neller 
1944) . The widespread drainage of the sawgrass 
Everglades lengthens the time during the year 
that these sediments are above water and could 
be expected to have an effect on the timing of 
occurrence of these different chemical forms of 
nitrogen. If conditions are adverse to some of 
the nitrifying bacteria, toxic levels of nitrite 
could accumulate. 

Tusneem and Patrick (1971) described nitrogen 
transformations in waterlogged soils used for 
jrowing rice in Louisiana. These conditions 
superficially resemble those where sawgrass grows 
laturally in the Everglades of southern Florida, 
Dut the soils are different in composition and 
ondition, the plants may well have different 
netabolic processes, and the climate is 
different . The rice soil that Tusneem and 
D atrick (1971) described, for example, was Crowley 
;ilt loam with a total nitrogen content of 770ppm. 
Iverglades muck contains from 3.0 to 3.8% total 
litrogen (Miller 1918, Hammer 1929) . Rice 
itilizes ammonium more efficiently than nitrate 

Department of Biology, University of Miami, 
loral Gables, Florida 33124. 

in all stages of growth (Bonner 1946) ; the 
nitrogen form required by sawgrass is unknown. 
The climate of Crowley, Louisiana, which lies at 
30°11' north latitude, is somewhat more temperate 
than that of the southern Everglades at 25°45' 
north latitude. 

The climate of southern Florida has been 
called "tropical savannah" (Hela 1952) . In 
general, it consists of a warm wet season that 
extends from May through September and a cool 
dry season that extends from October through 
April. Four seasons, however, can be identified 
as shown in Table 1. Winter, which lasts about 
three months from the first of December to about 
the end of February, is typified by mild 
temperatures, very little rainfall, falling 
water table, and minimum radiation due to low 
solar angle and short day length. Spring, last- 
ing from the first of March to the third week of 
May, is warmer than winter and also has very 
little rainfall. Spring is the driest of the 
seasons. The water table is lowest then and 
large areas of the surface of the Everglades, 
which had been submerged, are exposed. Radiation 
approaches the annual maximum. Summer begins in 
late May, or early June, with the onset of the 
seasonal rains. While daylength is maximum, 
radiation is not always maximum because of in- 
creased cloud cover. The water table rises 
rapidly when the rains begin and most of the 
Everglades basin becomes inundated. Summer, with 
its characteristically hot and humid conditions, 
prevails until about the middle of September. 
Daytime temperatures in fall can be as high as 
those in summer, but are usually lower, especially 
at night. Water levels in fall usually are 
initially high, but decline rapidly. Although 
days are shortening, with decreased cloud cover 
incident radiation can equal that measured in 
June. The mean temperature for January is 20°C 
and for August 28°C. The average rainfall is 
100 to 150cm, but annual and areal variations are 
great. For example, the Miami area had 123cm 
in 1975, but during the first 10 months of 1976, 
it had 220cm of rainfall. Summer rainfall is 
extremely variable in quantity and distribution. 
These vagaries can exert pronounced effects on 
local water levels. 

The climate is not the only force exerted on 
hydroperiod. Water levels in the Conservation 
Areas and water flow to Everglades National Park 
are controlled through impoundment and removal of 
water by pumping by the Flood Control District, a 
state agency. The canals, levees, and roads have 
altered the historical flow pattern of water over 
the surface of the Everglades basin. Instead of 
being inundated during the wet season and above 
water level during the dry season, there now are 
areas that are wetter for longer periods of the 
year {e.g., the southern ends of the Conservation 
Areas) and those that are drier for longer periods 
of the year (e.g., Taylor Slough and much of the 
eastern boundary of the Everglades) than before 
development . 

The nutrient requirements of economically 
important plants have been studied and reported 
at great length. Those of most plants growing in 


the natural state, including sawgrass, have not 
been reported (Gigon and Rorison 1972) . Steward 
and Ornes (1975) reported on mineral content of 
above ground parts of healthy sawgrass plants, 
the sediment, and surface water. They concluded 
that because sawgrass has low levels of mineral 
nutrients, especially phosphorus, it probably 
has low requirements for them and thus is able to 
dominate the Everglades ecosystem. 


The study site, L67, was chosen because it is 
in the middle of the Everglades basin, has 
vigorous sawgrass plants, and has shown the least 
change in overall appearance during the previous 
five years of study. It is in the north central 
part of the Shark River slough in an area free 
of tree islands. Nearby there are open glades 
in which either spike rush, maiden-cane, or cat- 
tail dominate. It is about 13km south of the 
Tamiami Trail and 200m west of Canal and Levee 67 
along the northeast edge of Everglades National 
Park. L67 has standing surface water longer in 
the dry season than do areas to the east and to 
the west of it; it did not dry down completely 
during 1975. It was burned by prescription on 
March 16, 1972, and previously by wildfire in 
March 1971. 

Samples fo 
at roughly mo 
were consider 
a community c 
follows: gre 
rhizomes, inf 
present) , top 
midlayer muck 
bottom muck ( 


r this study were taken during 1975 
nthly intervals. The plants and muck 
ed to be contiguous compartments of 
olumn. These compartments were as 
en leaves, brown leaves, roots and 
lorescences, surface water (when 
muck (0-10cm below muck surface) , 
(25-30cm below muck surface) , and 
50-70cm below muck surface) . 

Five entire mature plants representative of 
each site were carefully removed from the muck 
by hand. The roots were rinsed in the standing 
water on site to remove most of the adhering muck, 
and the intact plants were placed in plastic bags, 
which were then tightly sealed. In the laboratory, 
the plants were divided into the compartments 
listed above. These compartments were weighed, 
dried at 105°C in a forced-draft oven to constant 
weight, then weighed again. Plants were processed 
within 4 hours of harvest or stored frozen if 
processing was delayed. The dried compartments 
were ground in a Wiley mill to pass a 40-mesh 
screen. The samples were stored in tightly closed 
glass jars until analyzed. 

Muck samples were collected by hand and placed 
in tightly capped 12-ounce plastic cups. The 
muck samples included all of the water associated 
with the volume of muck taken. Water, when pres- 
ent above the surface of the muck, was collected 
in pyrex flasks that were closed with parafilm- 
wrapped rubber stoppers. Muck and water samples 
were placed on ice in the field immediately upon 
collection. If analyses were begun within 36 
hours, they were stored in a refrigerator at 4°C 
in the laboratory, otherwise they were stored 
frozen. Sediment samples were prepared for 
analyses by throughly mixing the fresh or thawed 
sample with a stainless steel spatula. Undried 
subsamples were used in all chemical analyses. 
Moisture and ash content were determined on sub- 
samples that were dried in a forced-draft oven at 
105°C and then ignited in a muffle furnace at 
550°C. Water samples were filtered through a 
washed 0.45um Millipore filter prior to storage or 
analysis . 

Total nitrogen and 
and muck samples were 
method, which includes 
described by Bremner ( 
converted to orthophos 
organic material with 
ignition at 400°C, as 
Orthophosphate was the 
chloride method after 
butanol to exclude sil 

total phosphorus in plant 
determined by a Kjeldahl 

nitrates and nitrites, as 
1965) . Phosphorus was 
phate by digestion of the 
Mg(N03)2 followed by 
described in AOAC (1971) . 
n determined by the stannous 
extracting it with benzene- 
icon (APHA 1971) . 

Water soluble components, i.e., ammonium, 
nitrate, nitrite, and phosphate, were extracted 
from subsamples of muck and plant material by 
suction filtration after shaking the material 
with deionized water. 

Soluble nitrite-nitrogen in the water extracts 
of plant and muck materials was measured by the 
modified Griess-Ilosvay method of Barnes and 
Folkhard (1951) . Soluble nitrate-nitrogen was 
reduced to nitrite by passage of an aliquot of 
each extract through a cadmium-copper column as 
described by Strickland and Parsons (1968) . Each 
aliquot was then analyzed for nitrite. Soluble 
ammonium-nitrogen was determined on each extract 
by the phenolhypochlorite method of So'lorzano 
(1969) . Soluble phosphate was determined as 
described above for orthophosphate. Water samples 
were analysed by the same methods used for water 
extracts. A Hatchi 139 spectrophotometer with 
lcm and 10cm cells was used. 

Known addition techniques, and positive and 
negative controls were routinely used to validate 
analytical method results. 

The following sensitivities were obtained with 
these methods based on one gram of oven-dried 
material, including ash: total Kjeldahl nitrogen, 
0.03mg/g; phosphorus, O.lppm; NO -N , 0.2ppm; 
N0 2 -N, 0.16ppm; NH 4 ~N, 0.023ppm. 


The results are shown in a series of bar graphs, 
Fig. 1, total phosphorus and soluble phosphate- 
phosphorus; and Fig. 2, total nitrogen, and the 
following water soluble components: ammonium- 
nitrogen, nitrate-nitrogen, and nitrite-nitrogen. 
The bars have varied widths to indicate that the 
sampling periods were not of uniform length as 
the four seasons are of unequal duration. Plant 
values are superimposed on sediment values for 
each component, which are given as mg/m^- 
community unit on a uniform log scale. A commu- 
nity unit is defined as an area one meter square 
to the depth of 30cm in the sediment, and all the 
plants growing in that space to whatever height 
they may attain. The sediment depth was limited 
to 30cm for uniformity. At this site the sediment 
was generally at least 30cm deep and even where it 
was deeper, the living sawgrass roots were 
primarily found in the upper 30cm of sediment. 
Biomass and density are given by Hofstetter and 
Parsons (1975) on the basis of a square meter, 
thus comparisons between these data can be made. 
Concentrations of nutrients in the surface water, 
i.e., the water above the surface of the sediment, 
are reported as parts per million in Table 2. 
The water contained very low concentrations of 
all components. 


TABLE 1. Climatic factors in the Everglades. Average values for 30 days prior 
to seasonal sampling date at L67, Everglades National Park. 


Dry Bulb 

Wet Bulb 


Day Length, 

Precip. , 

(Sample Date) 

Temp. ,C. 

Temp. ,C. 





































m 9/m J 

10 7 

io 6 H 
io 5 


io 3 H 

io 2 H 

io H 

io H 





c — r~T — i 



FIGURE 1. Quantity, mg/m-community unit, of total phosphorus and water-soluble 
phosphate contained in plants and the sediment (soil) at L67, Everglades National 
Park during 1975. 

TABLE 2. Concentrations of soluble components found in 
surface water at L67, Everglades National Park. 

































1 .-< 





] . i 




m 9/n 

10 7 -i 






10 5 - 

F- r ~ r ^~ 

io 4 - 

io 3 - 












10 1 

io 6 - 
io 5 - 
io 4 - 












FIGURE 2. Quantity, mg/nr'-community unit, of total nitrogen, and water-soluble 
ammonium, nitrate, and nitrite contained in plants and the sediments (soil) at 
L67, Everglades National Park during 1975. 

Total phosphorus was present in the sediment 
in quantities ten times greater than that found 
in plants. There was an appreciable increase in 
total phosphorus in plants in August primarily 
because there was an increase in biomass during 
summer, especially of roots, which had the 
greatest concentration of phosphorus . The in- 
crease in value of total phosphorus in plants 
from January to August was six-fold. 

Soluble phosphate was present in plants 
throughout the year and increased in quantity at 
about the same time (August) as total phosphorus. 
Its proportional increase during summer was 
greater; the soluble phosphate value reached 1/2 
that of the total phosphorus in plants and 
exceeded by a few milligrams that found in the 
sediment. Soluble phosphate in the sediment was 
found only in spring and summer. 

Total nitrogen in the sediment was present in 
almost 100 times the quantity found in plants 
throughout the year. There was a decrease in the 
quantity of total nitrogen during June, which 
coincides with the beginning of the wet season. 
This occurred in both the sediment and plants, 
but the sediment quantity still exceeded the 
quantity found in plants by 1000 times. The de- 
crease in soil nitrogen possibly could have been 
due to an increased denitrif ication activity. 
Losses of nitrogen and volatile nitrogen com- 
pounds directly from plants were studied by 
Pearsall and Billimoria (1937), Vickery et al . 

(1946) , Wilson (1943) , and Allison and Sterling 
(1948) . No attempt was made to account for this 
loss in this study. 

Ammonium was present in greatest quantities in 
winter in the sediment and also in the plants. 
This was during a period when water was fairly 
high, but was coolest and had the greatest amount 
of dissolved oxygen (e.g., at noon in January 
when water temperature was 26 °C, saturation would 
be 8mg/l, the surface water had 7mg/l) . 
Ammonium was present in greater quantities in 
April-May when water was shallow (7cm) than in 
September-October when water was at its maximum 
depth. The April-May value was about 100 times 
less than the winter value. Ammonium in plants 
was highest when ammonium in the sediment was 
highest. Ammonium increased again slightly in 
summer and except for one further increase in 
September, remained at a relatively uniform low 
level . 

Nitrate was present in greatest quantity when 
ammonium was not, both in the sediment and in 
plants. There was a decrease in the quantity of 
nitrate in plants in April-May. Hageman and 
Flescher (1960) found that nitrate in plant tops 
represents a surplus of uptake over reduction in 
roots or leaves. This decrease in nitrate in 
sawgrass at this time suggests an increase in 
reductase activity not seen at other times. 

Nitrite was found in the standing brown leaves 


when nitrate was absent, ammonium was high, and 
nitrite was not found in the sediment. This 
suggests a metabolic slowdown with residual ex- 
cess nitrate from the previous high-nitrate 
period caught in a transition state of reduction. 
The incidence of rather high amounts of nitrite 
in the sediment in fall also suggests a slow-down 
in the nitrifying process. 


The widely accepted view that submerged soils 
are anaerobic and contain nitrogen only as 
ammonium apparently does not apply to conditions 
at L67 in the sawgrass Everglades. Something 
Dther than the flooded condition determines the 
state of oxidation of nitrogen. Perhaps sawgrass 
aerates its substrate during its growing season 
sufficiently to allow aerobic nitrification dur- 
ing high water. In support of this conjecture, 
Forthman (1973) noted a growth surge in late 
spring, Conway (1937) demonstrated that Cladium 
mariscus, a related species, serates its roots 
by diffusion through the aerenchyma of mature 
green leaves and brown leaves. Dissolved oxygen 
was found at low but measurable levels in surface 
water even during summer at L67. 

Apparently there is an adequate supply of total 
phosphorus in the sediment to support the plants 
growing in it as it is always present in the sedi- 
ment in excess of that found in plants. Soluble 
phosphate, however, may be limiting at some times. 
The reason for the limited time of solubility and 
degree of dissolution is at present unknown. All 
of the soluble components were present as very 
small fractions of the total component. 

No clear relationship between the components 
in the sediment or in the plants was determined. 
Nitrate and ammonium were mutually exclusive, 
but did not follow the seasonal pattern with 
respect to each other that one would expect. 
'The components in sawgrass and the sediment seem 
to respond to some element of climatic seasonality. 
Hydroperiod seems to be the element of most 
importance, but the two major seasons coincide 
with differences in radiation, temperature, and 
increasing and decreasing day length. None of 
these seem to vary greatly during the year in 
southern Florida compared with values from more 
temperate latitudes, but without more information 
it is not possible to discount their importance. 


ALLISON, F. E. and L. D. STERLING. 1948. 

Gaseous losses of nitrogen from green plants. 
II. Studies with excised leaves in nutrient 
media. Plant Physiol. 23:601-608. 

AOAC. 1970. Methods of analysis of the AOAC, 
11th ed. Washington, D.C. 1015 pp. 

APHA. 1971. Standard methods for the examina- 
tion of water and wastewater. APHA and AWWA, 
Washington, D. C. 874 pp. 

BARNES, H. and A. R. FOLKHARD . 1951. The deter- 
mination of nitrites. Analyst 76:599-603. 

1954. Solution-culture studies of nitrite 
toxicity to plants. Soil Sci. Soc. Am. Proc . 

BONNER, J. 1946. The role of organic matter, 
especially manure, in the nutrition o rice. 
Bot. Gaz. 108:267-279. 

BREMNER, J. M. 1965. Total nitrogen. In: Meth- 
ods of soil analysis, Part 2. C. A. Black, 
(ed.), American So. of Agron., Madison, WI. 
1572 pp. 

CONWAY, V. M. 1937. Studies in the autecology 
of Cladium mariscus R. Br. III. The aeration 
of the subterranean parts of the plant. The 
New Phytol. 36:64-96. 

COURT, M. N., R. C. STEPHEN and J. S. WAID. 1962. 
Nitrite toxicity arising from the use of urea 
as a fertilizer. Nature (London) 194:1263- 

DAVY, A. J. and K. TAYLOR. 1974. Seasonal 

patterns of nitrogen availability in contrast- 
ing soils in the Chiltern Hills. J. Ecol. 

FROTHMAN, C. A. 1973. The effects of prescribed 
burning on sawgrass. Cladium jamaicense 
Crantz, in South Florida. MS Thesis. Univ. 
of Miami, Coral Gables, FL. 83 pp. 

GIGON, A. and I. H. RORISON. 1972. The response 
of some ecologically distinct plant species to 
nitrate- and to amonium-nitrogen. J. Ecol. 

HAGEMAN, R. H. and D. FLESCHER. 1960. Nitrate 
reductase activity in corn seedlings as affec- 
ted by light and nitrate content of nutrient 
media. Plant Physiol. 35:700-708. 

HELA, I. 1952. Remarks on the climate of South 
Florida. Bull, of Mar. Sci. of the Gulf of 
Mexico and Carribbean Sea. 2:438-447. 

HAMMAR, H. E. 1929. The chemical composition of 
Florida Everglades peat soils, with special 
reference to their inorganic constituents. 
Soil Sci. 28:1-13. 


The status of sawgrass, Cladium Jamaicense , in 
southern Florida. in .-Hof stetter , R. H. and 
Frances Parsons. Effects of fire in the eco- 
system. Final Report to the Dept. of the In- 
terior. EVER-N-48. 64 pp. 

HUTCHINSON, G. E. and V. T. BOWEN. 1950. Limno- 
logical studies in Connecticut. IX. A quan- 
titative radio-chemical study of the phos- 
phorus cycle in Linsley Pond. Ecology 31:194- 

MILLER, C. F. 1918. Inorganic composition of a 
peat and of the plant from which it was formed. 
J. Agr. Res. 13:605-609. 

NELLER, J. R. 1944. Influence of cropping, rain- 
fall, and water table on nitrates in Everglades 
peat. Soil Sci. 57:275-280. 

PEARSALL, W. H. and M. C. BILLIMORIA. 1937. 
Losses of nitrogen from green plants. Bio- 
chem. J. 31:1743-1750. 

POMEROY, L. R. 1960. Residence time of dis- 
solved phosphate in natural waters. Science 

RIGLER, F. H. 1956. A tracer study of the phos- 
phorus cycle in lake water. Ecology 37:550- 

S0L0RZAN0, L. 1969. Determination of ammonia in 
natural waters by the phenolhypochlorite 
methods. Limn, and Oceanogr. 14:799-801. 

STEWARD, K. K. and W. H. ORNES. 1975. The 
autecology of sawgrass in the Florida Ever- 
glades. Ecology 56:162-171. 

STRICKLAND, J. D. H. and T. R. PARSONS. 1968. 
A practical handbook of seawater analysis. 
Bull. 167. Fisheries Res. Bd . of Canada. 
Ottawa. 311 pp. 

TUSNEEM, M. E. and W. H. PATRICK, JR. 1971. 

Nitrogen transformations in waterlogged soil. 
Bull. No. 657, Dept. Agr. Agr. Expt. Sta., 
L. S. U., Baton Rouge, LA. 75 pp. 

WORTH. 1946. Chemical investigations of 
the metabolism of plants. I. The nitrogen 
nutrition of Narcissus poeticus . Conn. Agr. 
Expt. Sta. Bull. 406. 

VIETS, F. G., JR. 1965. The plant's need for 
and use of nitrogen in: Bartholomew, W. V. 
and F. E. Clark, eds., Soil nitrogen. 
Agronomy No. 10, Amer. Soc. of Agron., Madi- 
son, WI . 615 pp. 

WILSON, J. K. 1943. Gaseous losses of nitrogen 
from green plants. II. Studies with excised 
leaves in nutrient media. Plant Physiol. 



Richard Stalter 


Fire Island is a barrier island located along 
the south shore of Long Island extending from 
Southampton, L.I. to approximately the Nassau- 
Suffolk County line, a distance of 57 miles. 
It is separated from the mainland of Long Island 
by Great South Bay, which is from three-quarters to 
three miles wide. The island serves as a natural 
breakwater and is separated into three separate 
and distinct barriers by Shinnecock and Moriches 
Inlets . 

Fire Island is of rece 
developed sometime during 
earliest development invo 
submerged sand bar by the 
waves in the shallow wate 
shore. The steady accumu 
up by the breakers caused 
migrate landward eventual 
chain of islands that exi 

nt geological origin and 

the early Holocene. Its 
lved the formation of a 

breaking action of 
r some distance from 
lations of sand tossed 
the bar to emerge and 
ly developing into the 
sts today (Tourney 1959) . 

Other physical forces play a continuing role 
in the further development of the Fire Island 
barrier. Water movements, especially long-shore 
currents, are of prime importance in the main- 
tenance of the island. These currents result 
when water is brought into the near-shore area 
by waves that approach the island at a slight 
angle to the shore. The result is a small, 
net displacement of water and drifting sand, 
parallel to the shore. This process is un- 
doubtedly the most important source of sand 
deposition and erosion on the island today. The 
longshore drift runs predominantly from east to 
west and is responsible for the continued west- 
ward growth of the island, In the last 250 years 
historical records indicate a westward growth 
of almost six miles at Fire Island Inlet 
(Tourney 1959) . 

Beach drifting has also had its effects on the 
shaping of the barrier beach. The continual shift- 
ing of sand particles along the beach face by 
breaking waves results in the formation of a 
series of small ridges and valleys oriented 
parallel to the shore in the lower portions of the 
beach. These ridges and runnels have a seasonal 
development being partially obliterated during the 
winter only to be reformed during the summer. 
Deposition of dune sand on the landward side of 
the island is promoted as dry sand on the ocean side 
is blown landward by onshore winds. 


The climate of Long Island is mild. The 
average killing frost occurs in late October to 

The assistance of Professor William Nieter, who 
carefully reviewed and corrected the section deal- 
ing with geology is greatly appreciated. 

Environment Studies Program, St. John's Uni- 
versity, Jamaica, New York 11439. 

early November and can be expected from the 
middle of April to early May in the spring. The 
mean January temperature ranges from 30° in 
Medford to 3 2° in New York City, while the 
average July temperature ranges from 70° in 
Southampton to 77° in New York City. The average 
annual rainfall ranges from 43 inches to 45 
inches (U.S.D.A. 1941) . 


There are three general plant communities on 
Fire Island: the sand dune community, the salt 
marsh community, and an evergreen forest of 
Ilex opaca (holly) . 

The plants of the sand dune community 
generally tolerate salt spray. The most common 
members of this community include Ammophila 
breviligulata (dune grass), Lathyrus japonicus 
(beach pea), Artemesia stellariana (dusty 
Miller), and Solidago sempervirens (seaside 
goldenrod) . 

The salt marsh cpmmunity contains Spartina 
alterni flora (tall marsh cord grass), Salicornia 
spp. (glass wort) , Limonium carol inianum (sea 
lavender), Distichlis spicata (spike grass), 
Spartina patens (salt meadow grass) , and Iva 
oraria (high tide bush) . Soil salt content and 
tidal flooding account for the zonation of salt 
marsh species. 

The most unique plant community of the Sunken 
Forest occupies an area of approximately 64 
acres behind the second dune system. ilex 
opaca (holly) is the dominant species of the 
Sunken Forest which is one of the best developed 
evergreen maritime forests in the coastal north- 
eastern United States. 

At the present time, few workers have concerned 
themselves with the ecology and flora of Fire 
Island. Schulte (1965) prepared a floristic and 
ecological study of Fire Island, and commented on 
the parameters affecting the distribution of spe- 
cies in the Sunken Forest. Art et al . (1974) 
worked with the nutrient budget of the Sunken 
Forest. With the exception of individuals in- 
volved in floristic work, no investigator has 
attempted to study and characterize the three 
rmjor communities of the Fire Island National 

The Fire Island National Seashore, consisting 
of 19,311 acres of land, was established as a 
national seashore in 1964. At the present time, 
the park is visited by thousands of people each 
year, who walk the trails to examine the 
vegetation of the Sunken Forest. Surprisingly, 
members of the Torrey Botanical Club have not 
organized field trips to this unique area, and 
many botanists are unaware of this unique 



The present study consists of an examination 
of the parameters affecting the distribution of 
vegetation in three distinct plant communities. 
The vegetation of the sand dune community was 
examined by the quadrat method. One meter 2 
quadrats were established at five stations, 
including the front of the foredune, the top of 
the foredune, the depression behind the foredune, 
the top of the second series of dunes, and a 
depression behind the second series of dunes 
(Figure 1) . Soil samples were taken at the to 
2.5 and 15 centimeter levels and analyzed for 
salt content. Frequency data were prepared for the 
dune species and presented in Table 1. The salt 
content of the soil was measured following the 
procedure of Black et al . (1965). 

The salt marsh species were surveyed using a 
surveyor's transit and stadia pole, and the spe- 
cies arranged in increasing elevation above the 
species (Spartina al terni flora) Occupying the 
lowest elevation of salt marsh. Soil samples 
were taken and salinity values determined (Table 
2) . 

Since Schulte (1965) and Art et al. (1974) 
have studied the Sunken Forest community, the 
present investigation was restricted to sampling 
techniques not employed by these investigators. 

Ten 10 x 10 meter quadrats were established in 
the area of the Sunken Forest and frequency, 
density, basal area, and importance values calcu- 
lated for the arborescent species (Table 3) . in 
addition, size classes were established for the 
arborescent species of the forest (Table 4) . 
Soil texture data was measured for the Sunken 
Forest community (Table 5) . 

Classification of the plant species follows 
Fernald (1950) . 


Sand Dune Community 

The sand dune habitat is characterized by 
extremes of daily temperature, coarse porous 
soil, the affects of salt spray and drought. 

Frequency data suggest that dune grass was 
dominant in the first three study sites (Table 
1), with Solidago sempervirens (seaside 
goldenrod) , Artemesia stellariana (dusty Miller) , 
and Cakile edentula (sea rocket) important 
associates on the front of the ocean facing down 
where salt spray is heaviest (Oosting and Billings 
1942 and Stalter 1974) . Certain species that can 
tolerate lesser amounts of salt spray could be 
found just behind the top of the first dune. 
These included Prunus maritima (beach plum), 
Par thenoci ssus quinque folia (Virginia creeper) , 
while Hudsonia tomentosa (beach heather) was 
dominant at the depression behind the first dune. 
Hudsonia, a species that thrives in bright sun- 
light and shifting sand, cannot tolerate shade 
and high concentrations of salt spray and thus 
was only found in this habitat. A larger number 
of species was found en site 4 where the con- 
centration of salt spray is low, and on site 5 
where salt spray is minimal (Figure 1) . 

Studies on the east coast of the United 
States concerned with either the floristics or 
ecology of the sand dune community include Kurz 
(1942) and Davis (1942) for the Florida coast; 
Coker (1905) and Stalter (1974) for the coast 

of South Carolina; Kearney (1900) , Oosting and 
Billings (1942), Boyce (1954), and Burk (1962) 
for the coast of North Carolina; Chrysler (1911) 
for Maryland; Snow (1913) for Delaware; 
Harshberger (1900) for New Jersey; and Boyce 
(1954) for Massachusetts. 

Students of sand dune ecology have concurred 
with the findings of Oosting and Billings 
(1942) that salt spray and the transitory nature 
of the dunes (Boyce 1954, Stalter 1974) are most 
important in restricting sand dune vegetation. 

Salt Marsh Community 

The salient features and causes of the 
establishment and development of salt marshes in 
the northeastern United States have been reported 
by Chapman (1940, 1960) . Chapman reported that 
edaphic factors, such as chlorinity, salinity, and 
inter-species competition, coupled with duration 
and depth of flooding are important parameters 
that affect species distribution in salt marshes. 

Results in the present study indicate that the 
absence of certain salt marsh species in each zone 
of the marsh reflects the inability of that 
species to tolerate salinity (Kerwin 1966) ; 
inundation (Johnson and York 1915) , Chapman 1938) ; 
salinity and inundation (Chapman 1940) (Adams 
1963) ; or a complex of several factors (Miller 
and Egler 1950, Stalter 1968) . Spartina 
al terni flora (tall marsh cord grass) can tolerate 
the widest range of salinity and the longest 
period of flooding. The other species, Limonium 
carol i nianum (sea lavender) , Sal icornia v irg inica 
(salt wort), Suaeda linearis (Suaeda), Salicornia 

ovii (salt wort), Distichlis spicata (spike 
grass) , Spartina patens (salt meadow grass) , and 
Rtriplex patula (Atriplex) , while tolerating vary- 
ing ranges of salinity, occupy only a slight 
elevation gradient (usually a foot or less) in 
this marsh and in other marshes on Long Island 
(Lonergan and Stalter, unpublished data). 

iva oraria (high tide bush) occupies an area 
flooded infrequently at high tide. Soil 
salinities in this area are generally lower than 
soil salinities tolerated by other salt marsh 
species. Aster tenuifolius (salt marsh aster), 
while found associated with species in the upper 
portions of the salt marsh, is not as salt 
tolerant as the Spartinas, Limonium , or 
Salicornia. Baccharis halimifolia (silverling) 
can be found at the upper fringe of the salt 
marsh, but should not be classified as a strict 
salt marsh associate as thrives in open disturbed 
sites along Interstate 95 in North and South 
Carolina . 

Sunken Forest 

The data from the Sunken Forest indicates that 
ilex opaca is the dominant species. Other 
important associates include Amelanchier 
canadens i s (shad bush) , Sassafras albidum 
(sassafras); while two species of oaks, Quercus 
stellata (post oak) and Quercus velutina (black 
oak) , are unimportant associates . 

Many of the tree species are quite old. Cores 
taken from the holly on the island include one 
specimen with an age of 164 years while post 
oak, black oak, and juniper ( Juni perus virgi niana 
all are represented by trees over 100 years of 
age. Annual rings of one fallen oak were 
approximately 180 years. (Bullington 1976, 


TABLE 1. Frequency data for species found in five study areas at 
the Sunken Forest, Long Island. Station 1, the front of the first 
dune; station 2, the top of the first dune; station 3, the depres- 
sion behind the first dune; station 4, the top of the second dune; 
station 5, the depression behind the second dune (the Sunken Forest) 

Study Areas 















4 2 

J 3 













1 1 


















7 r » 


6 6 

7 5 












2 5 

2 5 








1. Ammophila brevil igulata 

2. Cakile edentula 

3. Sol idago sempervi rens 

4. Artemisia stellar iana 

5. Lathy rus japonicus 

6. Euphorbia pol ygoni fol ia 

1. Parthenocissus quinquefolia 

8. Prunus maritima 

9. Rhus radicans 

10. Hudsonia tomentosa 

11. Arctostaphylos uva-urs i 

12. Ilex opaca 

13. Lactuca Sp . 

14. Smilax rotundifolia 

15. Achi 1 1 ia millifolium 

16. Rosa rugosa 

17. Prunus serotina 

18. Amelanchier stel lata 

19. Smilacena stel lata 

20. Rhus copal 1 ina 

21. Rubus sp. 

22. Ma ianthemum canadense 

23. Juniperus virg iniana 

24. Myrica pennsy lvanica 

25. Smilax glauca 

26. Vacc inium corymbosum 

27. Rubus hispidus 

28. Pter idium squil inum 

29. vitis sp. 

30. Sassafras albidum 

31. ARalia nudicaul is 

32. Ribes cynosbati 

33. Lonicera japonica 

34. Viburnum dentatum 

TABLE 2. A list of salt marsh psecies in the 
Fire Island National Seashore. 


Salinity range 

Spartina al terni flora 
Sal icornia virginica 
Salicornia bigelovii 
Limonium carol inianum 
Suaeda 1 inearis 
Distichl is spicata 
Spartina patens 
Atriplex patula 
Iva oraria 
Bacchar is hal imifol ia 






TABLE 3. Density (D) , relative density (RD) , frequency (F), 
relative frequency (RF) , basal area (BA) , relative dominance 
(RD) , and importance value of arborescent species of the 
Sunken Forest, Fire Island, New York. 


Ilex opaca 
Sassafras albidum 
Amelanchier canadensis 
Nyssa syl vatica 
Quercus stellata 
Q. velutina 




6 / 








I V 

57 156 








1. 3 






4 6 






1 1 

3 2 






















FIGURE 1. A diagram of the dune, shrub, and forest plants of Fire Island in the area 
of the Sunken Forest. Site I (front of first dune facing ocean), site II (top of 
forest dune) , site III (depression behind first dune) , site IV (top of second dune) , 
Site V (depression behind second dune--the Sunken Forest) . 



















Successional trends on Fire Island in the Sunken Forest, New York. 

TABLE 4. Distribution of trees in four size classes (expressed 
as a percent) of the Sunken Forest, Fire Island, New York. 

Size class 










Ilex opaca 





Sassafras albidum 


1 2 



Amelanchier canadensis 

2 4 

J 4 



Nyssa sylvatica 





Quercus stel lata 





Q . vel uti na 






TABLE 5. Fire Island soil texture 

Sieve dimensions Classification % of soil fraction 

3. mm 
1 . 5 mm 
. 75 mm 
0.4 mm 
0. 25 mm 

Fine gravel 
Fine gravel 
Coarse sand 
Medium sand 
Fine sand 


pers. comm. ) . 

The findings of the present study are in accord 
with Schulte's (1965) survey of Fire Island. The 
Sunken Forest, because it lies behind a secondary 
stable dune system, is completely protected from 
salt spray and wind. The October hurricane of 
19 38 that ravaged the eastern end of Long Island 
had no serious effect upon the plant species of 
the Sunken Forest (Bullington 1976, pers. comm.). 
Schulte contends that birds probably brought the 
original holly seeds to the island from the main- 
land. The copious development of plants whose 
fruits are favored by birds lends credence to 
Schulte's observations. 

Schulte contends that there is a paucity of 
holly saplings. Sassafras and Nyssa sylvatica , 
that are well represented by saplings, may even- 
tually assume greater importance in the forest 
canopy. An examination of size classes of the 
trees of the Sunken Forest (Table 4) might negate 
Schulte's observation, since many of the holly 
trees may be over 100 years of age. 

Holly is well represented in all size classes. 
If one assumes that the smaller holly trees are 
younger than the larger hollies (an assumption 
that is not always a correct one) then it is prob- 
able that holly may maintain its dominance in the 
overstory . 

At Island Beach State Park, a barrier island 
similar to Fire Island, back dune succession leads 
to an association of Pinus rigida (pitch pine), 
Juniperus virginiana (red cedar), and Prunus sero- 
tina (black cherry) (Martin 1959) . Schulte ob- 
served several large dead specimens of Pinus rigida 
on Fire Island and maintains that Pinus rigida may 
have been more important in the Sunken Forest in 
the past. Pinus rigida, like Ilex, will grow in 
acid soil (pH 4.5 to 6) (Schulte 1965) but unlike 
ilex, Pinus is intolerant to shade. Therefore, if 
Pinus rigida were important on Fire Island, it may 
have been gradually shaded out. 

On Fire Island, a 
evolve into a commun 
canadensi s which in 
ity dominated by He 
sylvatica and Acer r 
Figure 2 depicts sue 
on Fire Island in th 
The data in this stu 
that the location of 
of holly, the presen 
and the presence of 
the dominance of He 

mixed shrub community may 
ity dominated by Amelanchier 
turn may develop into a commun- 
x. On the wettest sites, Nyssa 
ubrum (red maple) can be found, 
cessional trends that may occur 
e area of the Sunken Forest, 
dy and by other workers suggest 

the island, the shade tolerance 
ce of a high complex dune system, 
salt spray probably account for 


ADAMS, D. A. 1963. Factors Influencinj Vascular 
Plant Zonation in North Carolina Salt Marshes. 
Eco. 44:445-456. 

ART, H. W., F. H. BORMANN, G. K. BOIGHT, and 

G. M. WOODWELL. 1974. Barrier Island Forest 
Ecosystem: Role of Meteorologic Nutrient 
Inputs. Science 184:60-62. 

BLACK, C. S., D. D. EVANS, J. J. WHITE, L. E. 

ENSMINGER, and F. E. CLARK. 1965. Methods of 
Soil Analysis. Agronomy. No. 9. Part 1 and 
2, 1572 pp. 

BOYCE, S. G. 1954. The Salt Spray Community. 

Ecol. Monogr. 24:26-67. 
BURK, C. J. 1962. The North Carolina Outer 
Banks: A Floristic interpretation. Jour. 
Elisha Mitchell Sci. Soc . 78(l):21-28. 
CHAPMAN, V. J. 1938. Studies in Salt Marsh 
Ecology; Section I to III. Jour. Ecol. 26: 

. 1940. Studies in Salt Marsh Ecology. 

Sections VI and VII. Comparisons with Marshes 
on the East Coast of No. America Jour. Ecol. 

. 1960. Salt Marshes and Salt Deserts 

of the World. The University Press, Aberdeen, 
392 pp. 
CHRYSLER, M. A. 1911. Ecological Plant Geography 

of Maryland. Md. Weather Serv. 3:148-197. 
COKER, W. C. 1905. Observations on the Flora 
of the Isle of Palms, Charleston, S. C. 
Torreya 5:135-45. 
DAVIS, J. H. 194 2. The Ecology of the Vegeta- 
tion and Topography of the sand Keys of 
Florida. Tortugas Lab. (Carnegie Inst. Wash.) 
Paps. 33:114-195. 
FERNALD, M. L. 1950. Gray's Manual of Botany, 

8th ed. New York: American Book Co. 
HARSHBERGER, J. W. 1900. An Ecological Study 
of the New Jersey Strand Flora. Acad. Nat. 
Sci. Phila. Proc. 52:623-671. 
JOHNSON, D. S. and H. H. YORK. 1915. Relation 
of Plants to Tide Levels. Carnegie Institu- 
tion of Wash. Publication No. 206. 
KEARNEY, T. H. 1900. The Plant Covering of 
Ocracoke Island; A Study in the Ecology of 
the North Carolina Strand Vegetation. U.S. 
Natl. Herbarium. Contr . 5:261-319. 
KERWIN, J. A. 1966. Classification and Struc- 
ture of the Tidal Marshes of the Poropotank 
River, Virginia. ASB Bulletin 13:40. Abstract. 
KURZ. H. 1942. Florida Dunes and Scrub, Vegeta- 
tion and Geology. Fla. Geol. Surv. Bull. 23. 
154 pp. 
MARTIN, W. E. 1959. The Vegetation of Island 
Beach State Park, New Jersey. Ecological Mon- 
ographs 29:1-46. 
MILLER, W. R. and F. E. EGLER. 1950. Vegetation 
of the Wequetequock-Pawcatuck Tidal Marshes, 
Connecticut. Ecol. Monogr. 20:143-172. 
OOSTING, H. J. and W. D. BILLINGS. 1942. Fac- 
tors Effecting Vegetation Zonation on Coastal 
Dunes. Ecology 23:131-42. 
SCHULTE, E. 1965. A Study of the Plants in 
the Sunken Forest, Fire Island, New York. 
Unpublished Masters' Thesis. C. W. Post 
College, Long Island University. 
SNOW, L. M. 1913. Progressive and Retrogres- 
sive Changes in the Plant Associations of 
the Delaware Coast. Bot. Gaz. 55:45-55. 
STALTER, R. 1968. An Ecological Study of a 

South Carolina Salt Marsh. Unpublished Ph.D. 
Dissertation. 63 pp. 

. 1974. Vegetation in Coastal Dunes 

of South Carolina. Castanea 39:95-103. 
TOUMEY, D. 1959. Fire Island: Map 1798. Long 

Island Forum 22: 47-48. 
U. S. DEPT. OF AGRICULTURE. 1941. Climate and 
Man. United States Govt. Printing Office, 
Wash. D. C. 




Edward E. Dale, Jr. and James W. Gibbons 


Pea Ridge National Military Park is located in 
the western part of the Oak-Hickory Association 
(Braun, 1950) . This association, particularly in 
northwest Arkansas, is characterized by patches 
of prairie grasslands interspersed with forest 
(Turner, 1935) . 

Since such grasslands were reported by Bussey 
(1862) as present during the Civil War in what is 
now the western part of the park, a project was 
initiated in 1975 to re-establish prairie vegeta- 
tion in this area. 

Prairie restoration has been accomplished 
successfully in several areas of the United States 
including Iowa (Anderson, 1946; Christiansen, 1967), 
Illinois (Schramm, 1970), Wisconsin (Ode, 1970) 
and in Missouri (Toney, 1975), but such a project 
has not been previously attempted in Arkansas. 

The primary objective of this project was to 
re-establish a prairie at Pea Ridge National Mili- 
tary Park that resembles original prairies of this 
area as closely as possible. A secondary objec- 
tive was to gain experience in prairie restoration 
techniques so that expansion of the prairie area 
in the park could be facilitated later. 


A one-acre (.405 hectares) plot was selected 
approximately 2.2 miles northwest of park head- 
quarters as the site of the proposed prairie in 
April, 1975, and soil samples were taken for 
analysis by the University of Arkansas soils test- 
ing and research laboratory. The plot was plowed 
and disked twice, once in May and again in June 
to provide better soil conditions for the prairie 
species and to help eliminate weeds. 

Results of the soil analysis indicated that 
soil conditions at the Pea Ridge site were essen- 
tially similar to tall grass prairie sites in 
northwestern Illinois (Alexander, Fehrenbacker 
and Ray, 1968) and to places in northwest Arkansas 
that supported prairie species (Phillips, 1967) 
except that the soil was characterized by a low 
phosphorus and potassium content. Accordingly, 
23 pounds (20.4 kg.) of phosphorus (P2O5) and 30 
pounds (13.6 kg.) of potassium (K20) fertilizer 
were applied. 

Nitrogen was not added because the level present 
was considered adequate for the first year. Also, 
according to Owensby (1970) excess nitrogen would 
increase the chance of ecesis by cool season 
grasses which were abundant near the site. 

The area was divided into sixteen square sub- 
plots fifty feet (15.24 meters) on each side and 
marked by stakes. Also, a 1.5 meter wile strip 
extending the entire length of the site on the 
east side was designated as a separate experimental 

■""Department of Botany and Bacteriology, Univer- 
sity of Arkansas, Fayetteville 72701. 

Ecotypes of five species of prairie grasses 
(designated as horticultural varieties) known to 
be dominant or important constituents of prairies 
in northwest Arkansas and adjacent areas of 
Oklahoma and Missouri were chosen for initial re- 
establishment. The species and varieties used and 
seeding rates as shown in Table 1 were based on 
recommendations made by Roundtree, Bjugstad and 
Wheaton (1970) for Missouri, and by Stoin (1975) 
who has worked on pasture improvement in Arkansas . 

On June 13, 1975, all seeds except Panicum 
virgatum were mixed and hand broadcast over the en- 
tire restoration area except in the experimental 
plot and in a few selected areas. p. virgatum was 
planted in pure stands only in the more mesic areas 
of the site because it is a very aggressive compe- 
titor in moist area and was observed by the senior 
author to increase greatly relative to other spe- 
cies in tall grass prairie in Oklahoma over a five- 
year period. Seeds of Bouteloua curtipendula, 
Andropogon scoparius , Sorghastrum nutans, and 
Andropogon gerardi respectively, were planted in 
pure stands on an increasing soil moisture gradient 
at obviously more dry or wet sites in accordance 
with the classification of prairie species on a 
moisture continuum in Wisconsin as described by 
Curtis (1955) . 

According to Schramm (1970) seeds should be 
raked in and then rolled or trampled after plant- 
ing to increase germination. Ode (1970) states 
that mulch is significantly better than no mulch 
in establishing native grasses in Wisconsin using 
Nebraska seeds. Accordingly, the seeded plot was 
hand-raked and trampled, and fifteen bales of oat 
straw were evenly distributed as mulch. 

Since these procedures were suggested for other 
areas where soil and climate conditions are dif- 
ferent from those in Arkansas and different eco- 
types were utilized, it seemed desirable to make 
a quantitative study of different seedbed treat- 
ments near the restoration site to test these 
procedures for this area. 

The experimental plot area on the east side of 
the restoration site was divided into thirty-two 
1.5 by .75 meter sub-plots, and one hundred seeds 
of each of the five grasses were broadcast June 
20, 1975 in each sub-plot and submitted to all 
combinations of raking, mulching, and trampling 
or no treatment with four replications. After 
two months of growth, the number of seedlings 
present, by species, per treatment was counted 
and a two-way analysis of variance (Snedecor, 
1956) was used to determine significance of re- 
sults of the plot studies to the .05 level. 

Percent cover of living prairie and weedy spe- 
cies present was determined in September, 1975, 
and periodically during the growing season of 1976 
using a modification of the point method as de- 
scribed by Levy and Madden (1933). 

Sods of prairie grasses and forbs were collect- 
ed from an area near the park and transplanted to 
the prairie early in July, but further transplant- 


of pure 
Military Park. 


Species and varieties of prairie grasses and pounds per acre 
live seeds (PLS) planted in prairie site at Pea Ridge National 

Seeds planted 
Variety lbs PLS per acre 

Andropogon gerardi Vitman "Kaw" 6 (2.7 kg) 

A. scoparius Michx. "Aldous" 5 (2.3 kg) 

Sorghastrum nutans (L.) Nash "Cheyenne" 3 (1.4 kg) 

Bouteloua curtipendula (Michx.) Torr. "El Reno" 1.5 (.68 kg) 

Panicum virgatum L. "Blackwell 1 (.45 kg) 

ing was terminated because of drought conditions 
in late July and August. Also, observations of 
survival of transplants and prairie grass seedlings 
and counts of prairie forbs that became established 
naturally were made throughout 1975 and in 1976. 

Establishment of another acre of prairie adja- 
cent to the restoration plot established in 1975 
was initiated on June 7, 1976. 

The procedures followe 
plot were similar to thos 
ceptions based on experie 
first year's work and on 
(1976) . The amount of se 
approximately twice that 
except Andropogon scopari 
mately three times as muc 
done with the hope of obt 
ment in 1976. Also, raki 
ment given the 1976 resto 
following seeding, and fe 
July instead of June. 

d in establishing the new 
e used in 1975 with ex- 
nce gained during the 
a study made by Gibbons 
eds planted in 1976 was 
of 1975 for all species 
us, in which approxi- 
h was used. This was 
aining better establishi- 
ng was the only treat- 
ration site immediately 
rtilizer was applied in 

Photographs were taken and voucher specimens 
collected periodically during 1975 and 1976. All 
specimens are on file in the herbarium at the 
University of Arkansas, Fayetteville . 



Germination of grass seeds was noted approxi- 
mately ten days after planting. Actual percentage 
of germination was obscured by heavy rains that 
fell approximately one week after planting when 
large numbers of seeds were washed from areas ori- 
ginally seeded. However, each tenth-square meter, 
except those heavily washed, contained at least 
one and usually several seedlings. 

Seedlings had attained an average height of ap- 
proximately 7.6 cm by mid-July, and by mid-August, 
the grasses were an average of 20 cm tall and had 
begun to tiller and form sods. The average height 
of the grasses varied from 40.9 cm in p. virgatum 
to 20.2 cm in A . scoparius on September 23. 

First-year establishment of these grasses at 
Pea Ridge in 1975 resulted in plentiful flower and 
seed head production in August by B. curtipendula 
and isolated flowering in all others except p. 
virgatum. According to Schulenberg (1970) flower- 
ing during the first year was an indication of 
successful establishment in a restoration plot in 
Illinois . 

TABLE 2. Percent cover of prairie grasses, by species, total percent cover of 
other species collectively and percent of bare ground or dead vegetation on 
different dates in plots established in 1975 and 1976. Agi, Andropogon gerardi: 
Asc, A . scoparius ; Bcu, Bouteioua curtipendula; Pvi , Panicum virgatum; and Snu, 
Sorghas trum nutans. 

Established June, 1975 


grasses 9/7/75 5/20/76 7/8/76 9/7/76 10/7/76 

Established June, 1976 
9 7 76 10/13 7 6 
















3 .2 



Pv l 


. 1 

! .6 

3. 5 

1 .5 















62. 5 












or dead 
















; i 



2. 1 










Percentage cover of the five species of prairie 
grasses on September 7 ranged from 7.0% for a . 
gerardi to 0.6% each for B. curtipendula and P. 
virgatum. The five species covered a total of 
17.0% (Table 2) . 

Some weeds were always present at 
prairie site, but they were not parti 
able in large numbers until mid-Augus 
time Mullugo verticillata and Diodia 
ered much of the ground on bare areas 
ber 7, the important weedy species we 
] verticillata, Digitaria sanguinalis, 
\ Ambrosia a rtemis ii fol i a , Echinochloa 
' and Bidens aristosa (Table 3) . These 
: covered a total of 42.5 percent cover 
time the five prairie grasses covered 
fifth of the total area. 

the Pea Ridge 
cuiarly notice- 
t . At that 
teres cov- 

By Septem- 
re M . 
D . teres , 
crusgalli , 

species alone 
At this 

less than a 

TABLE 3. Vegetation of prairie site established 
| in June 1975 at Pea Ridge National Military Park. 
, September 7, 1975. 



Mollugo verticillata L. 

Andropogon gerardi Vitman* 

Digitaria sanguinalis (L.) Scop. 

Diodia teres Hall. 

Ambrosia artemisiifolia (L.) Des . 

Echinochloa crusgalli (L.) Beauv. 

Sorghastrum nutans (L.) Nash.* 

Bidens aristosa (Michx. ) Britt. 

Andropogon scoparius Michx.* 

Gaura biennis L. 

Tridens flavus (L.) Smyth 

Leptoloma cognatum (Schult.) Chase 

Setaria glauca (L. ) Beauv. 

Euphorbia corollata L. 

Panicum anceps Michx. 

Solanum carolinense L. 

Bouteloua curtipendula (Michx.) Torr.* 

Panicum virgatum L.* 

Rubus sp. 

Cyperus globulosus Aubl. 

Oenothera laciniata Hill 

Apocunum cannabinum L. 

Aristida oligantha Michx. 

Desmodium sp. 

Croton capi tatus Michx. 

Bare ground or dead vegetation 







*Denotes prairie grasses established by seeds. 

Results from the experimental plot studies 
showed that significant differences at the 0.05 
level were found between the number of seedlings 
surviving after two months. 

Raking appears to be the most effective treat- 
ment for seedling establishment. None of the 
other treatments were significantly better for 
seedling survival except for Sorghastrum nutans 
which showed significantly better results when 
mulching was applied with raking. 

Mulching only was of no benefit to seedling 
establishment, probably due to adequate soil mois- 
ture conditions at the site following planting. 
Sixty percent of the average yearly rainfall at 
Pea Ridge falls during the period from April to 
September (United States Department of Commerce, 
1974) , and 1975 was a particularly wet year. 
Hence, the moisture requirements of these species 
were met in 1975 and mulching to conserve soil 
moisture was unnecessary. However, during years 
with low rainfall, mulching might be required. 

Trampling was also found to be of no benefit to 
establishment at Pea Ridge. After rains, the 
soil at the restoration site tended to become com- 
pacted because of its high clay content. This 
possibly had the same effect on seed germination 
as trampling would have on coarser textured soils. 

Transplant experiments resulted in varying de- 
grees of success in establishing prairie species 
at Pea Ridge. Sods of the five seeded species, 
Elymus canadensis L. and Tripsacum dactyloides L. 
that were transplanted into the area in July were 
living in September 1975. 

Attempts were made to transplant ten species 
of prairie forbs. These transplants appeared at 
first to be successful, but drought conditions and 
excess browsing by white-tail deer eliminated all 
but Asclepi as tuberosa L. and Dalea purpurea Vent, 
by fall, 1975. 

It is probable that transplanting may be the 
best method for establishing species in a re- 
establishment prairie (Christensen , 1967; Bland, 
1970; Schulenberg, 1970); however, the technique 
requires so much time and hand labor that the ex- 
pense may be prohibitive. Also, sufficient a- 
mounts of transplant materials are frequently 
difficult to find. 

An unexpected benefit to establishment of 
prairie forb species occurred through natural 
growth from roots, rhizomes, or seeds. Prairie 
species that became established naturally in- 
cluded Vernonia missourica Raf., Euphorbia 
corollata L. , Solidago altissima L. , Apocynum 
cannabinum L., Schrankia Nuttallii (DC.) Stand L. , 
Rudbeckia hirta L. and Erigeron str igosus Muhl . 



In the discussion that follows, the prairie 
restoration area established in 1976 will be re- 
ferred to as the one-year plot and the area 
established in 1975 (which has supported prairie 
vegetation for two growing seasons) will be desig- 
nated as the two-year plot. 

One-Year Plot Established in 1976 

Seeds planted on June 7, 1976 did not germinate 
until the last week in June, primarily because of 
lack of moisture. However, a good stand of 
grass was evident over most of the area by July 8 
following a heavy rain in late June. 

Seedlings averaged approximately 26 cm high by 
mid- July and 10.4 cm by mid-August in all parts of 
the area except in places where fertilizer was 
applied in June. Average height of the grasses on 
the fertilized areas was 13.0 cm in mid-August, in- 
dicating that use of fertilizer on newly established 
prairie sites promotes better growth and faster es- 
tablishment of prairie species. This is not in 
agreement with recommendations of Schramm (1970) 
who advises against fertilizer application on new 
prairie restoration sites. Growth in the initially 
unfertilized areas became more rapid after ferti- 
lizer was applied in July. Prairie grasses aver- 
aged 15.6 cm in height in both fertilized and un- 
fertilized areas on October 7 indicating that 
mid-summer fertilizer applications can be 

Percent living cover of all prairie species was 
24.5 on September 7, but dropped to 19.4 by Octo- 
ber 13. However, the total cover of all other 
species combined (mostly weeds) increased from 40.5 
to 47.0 percent (Table 2). This is not surprising 
since prairie species normally undergo senescence 


and annual weeds frequently grow relatively faster 
at this time of year. 

The most common weed species present was Setaria 
glauca which comprised 11 percent cover. Paspalum 
laeve was second with a percent cover of 4.5, and 
all others present were 2.5 percent or less (Table 
4) . 

TABLE 4. Vegetation of prairie site established 
in 1976 at Pea Ridge National Military Park. 
September 7, 1976. 



(Michx.) Torr.* 

Setaria glauca (L.) Beauv. 

Andropogon gerardi Vitman* 

Echinochloa crusgalli (L.) Beauv. 

Sorghastrum nutans (L.) Nash.* 

Andropogon scoparius Michx.* 

Paspalum laeve Michx. 

Bouteloua curtipendula 

Panicum virgatum L . * 

Gaura biennis L. 

Mollugo verticillata L. 

Digitaria sanguinal is (L.) Scop. 

Panicum anceps Michx. 

Leptoloma cognatum (Schult.) Chase 

Ambrosia artemisiifolia (L.) Des. 

Diodia terres Hall. 

Croton sapitatus Michx. 

Bidens aristosa (Michx.) Britt. 

Solanum carol inense L. 

Aristida longespica Poir. 

Panicum lanug inosum Ell. 

Tridens flavus (L.) Smyth 

Panicum laxif lorum Lam. 

Tr i folium reflexum L. 

Bare ground or dead vegetation 






♦Denotes prairie grasses established by seeds. 

Naturally occurring prairie forbs present in 
the plot in September were Solidago altissima and 
Gnaphalium obtusi folium L. Others observed in the 
area during June and July that had completed their 
growth and reproductive cycles before September 
included Rudbeckia hirta , Euphorbia corol lata , 
Apocynum cannabinum and Shrankia Nut tail ii . 

Composition of First Year's Growth in 
1975 and in 1976 

Several differences were noted in species pres- 
ent and growth condition during the first year's 
growth in the two-year plot established in 1975 
and the one-year plot established in 1976. 

Germination and growth began earlier in 1975 
than in 1976, and average sizes of all prairie 
species were larger at the end of the growing sea- 
son in 1975. This is attributed primarily to more 
favorable moisture conditions. 

A heavy rain occurred the day after seeds had 
been planted on June 13, 1975, but it did not rain 
until nearly three weeks after planting on June 7, 
1976. Also, rainfall was greater and more evenly 
distributed during the 1975 growing season. 

Percent cover of prairie grasses was 17 percent 
on September 7, 1975 in the two-year plot and 24.5 
percent in the one-year plot on September 7, 1976 
(Table 2) . Apparently the greater seeding rate 
must have favored the higher survival percentage 
and compensated somewhat for the poorer growth 
conditions . 

It is noteworthy that weed populations were 
different near the end of the growing season of 
both years. Principal species present in 1975 
were Mollugo verticillata, Digitaria sanguinal is , 
Diodia teres, Ambrosia artemisiifolia, Echinochloa 
crusgalli and Bidens aristosa. In 1976, the prin- 
cipal species were Setaria glauca, Echinochloa 
crusgalli, and Paspalum laeve. Also, there was 
less bare ground in 1975 than in 1976 (Tables 3 
and 4 ) . 

Two-year Plot Established in 1975 

Increased growth and spreading of prairie 
species during the second year was noted on May 20 
1976, when cover of prairie species had increased 
to 30.2 percent. Cover percentage was 33.6 on 
July 8 and 62.5 on September 7, but it declined tc 
44.0 percent on October 7 near the end of the 
growing season (Table 2) . 

The general trends of increasing cover percent' 
ages from May to September and decline in October 
occurred in all prairie species except Sorghastrum 
nutans which had the highest percent cover in 
October, and in Bouteloua curtipendula which had ; 
cover percentage of 8 percent on May 20. 

Total percent cover of all other species 
collectively declined from May until October, 
probably because of competition with the increas- 
ing prairie grasses. The percentage of bare 
areas or dead vegetation declined from 25.5 per- 
cent on May 20 to 5 percent on September 7. It 
increased to 28.9 percent on October 7 primarily 
because of natural senescence. 

Populations of weed species changed considera- 
bly during the growing season with Aristida 
longispica , Tri folium reflexum , and Bidens 
aristosa occurring as principal species in Sep- 
tember (Table 5) . 

TABLE 5. Vegetation of prairie site established 
1975 at Pea Ridge Natioanl Military Park. 
September 7, 1976. 


Perce . 
cove ' 

Sorghastrum nutans (L.) Nash.* 

Andropogon scoparius Michx. 

Andropogon gerardi Vitman* 

Bouteloua curtipendula (Michx.) Torr. 

Aristida longespica Poir. 

Trifolium reflexum L. 

Bidens aristosa (Michx.) Britt. 

Panicum virgatum L.* 

Gerardia fasciculata Ell. 

Cassia fasciculata Michx. 

Solanum carol inense L. 

Panicum lanug inosum Ell. 

Tridens flavus (L.) Smyth 

Apocynum cannabinum L. 

Ambrosia artemisiifolia (L.) Des. 

SEtaria glauca (L.) Beauv. 

Paspalum -aeve Michx. 

Rubus sp. 

Euphorbia corol lata L. 

Diodia teres Hall 

Eupator ium serotinum Michx. 

Solidago altissima L. 

Bare ground or dead vegetation 











♦Denotes prairie grasses established by seeds 


It was noted that many weeds present in both 
1975 and 1976 were generally much smaller and less 
numerous during the second year. This was par- 
ticularly true of Ambrosia artemisii folia and 
Bidens aristosa which were very conspicuous in 

Prairie forbs transplanted into the plot in 

1975 that appeared to be established in August, 

1976 were Asclepias tuberosa, Dalea purpurea and 
Echinacea pallida Nutt. 

All species of naturally occurring prairie forbs 
present in 1975 were present in 1976 also. New 
species noted in 1976 included Gerardia fasciculata 
and Gnaphalium obtusifolium. 


It is concluded that satisfactory progress is 
being made toward the establishment of a prairie 
in both the one- and two-year plots using the tech- 
niques presently employed. This is based on the 
great increase in prairie species relative to 
weeds at the end of two years and the relatively 
higher percentage of prairie species present in the 
one-year plot established in 1976 as compared to 
first year's growth in 1975. Also, it is encour- 
aging to note that a comparison of vegetation of 
the two-year plot with a native mixed lowland 
prairie near Harrison, Arkansas, shows that per- 
cent cover and species composition of grasses in 
these two areas are similar. The forb population 
on the two-year plot lacks many species present 
in the native prairie and the weed population is 
still too high. In addition, there are several 
bare places that will require reseeding of prairie 
grasses and more intensive work on establishment 
of forbs is needed. However, indications are that 
with proper management most of the difficulties 
described can be greatly reduced or eliminated. 


RAY. 1970. Characteristics of dark colored 
soils developed under prairie in a topose- 
quence in northwestern Illinois. In: Schramm, 
Peter (ed.). Proceedings of a symposium on 
prairie and prairie restoration. Knox College 
Biol. Field Stat. Special Publ. No. 3. 

ANDERSON, W. A. 1946. Development of prairie 
at Iowa Lakeside Laboratory. Amer . Mid. Nat. 

BLAND, MARILYN K. 1970. Prairie establishment 
at the Michigan Botanical Gardens. in: 
chramm, Peter (ed.). Proceedings of a sympo- 
sium on prairie and prairie restoration. Knox 
College Biol. Field Stat. Special Publ. No. 3. 

BRAUN, E. LUCY. 1950. Deciduous forests of 

eastern North America. Blakiston Co., Phila- 
delphia. 596 pp. 

BUSSEY, COL. CYRUS. 1862. Description of vege- 
tation during the Battle of Pea Ridge. O. R. , 
Ser. 1, Vol. VIII:233. 

CHRISTIANSEN, PAUL A. 1967. Establishment of 
prairie species in Iowa by seeding and trans- 
planting. Ph.D. Thesis. Iowa State Univ., 
Ames, Iowa. 

CURTIS, J. T. 1955. A prairie continuum in Wis- 
consin. Ecology 36:558-566. 

GIBBONS, JAMES W. 1976. Reestablishment of 

prairie vegetation at Pea Ridge National Mili- 
tary Park, Benton County, Arkansas. Master's 
Thesis. University of Arkansas, Fayetville. 

LEVY, E. B., and E. A. MADDEN. 1933. The point 
frame method of pasture analysis. New Zealand 
Jour. Agri. 46:1933. 

ODE, ARTHUR H. 1970. Some aspects of establish- 
ing prairie species by direct seeding. in: 
Schramm, Peter (ed.) 1970. Proceedings of a 
symposium on prairie and prairie restoration. 
Knox College Biological Field Station Publica- 
tion No. 3. Galesburg, Illinois. 

SON. 1970. Effects of clipping and supple- 
mental nitrogen and water on loamy upland 
bluestem range. J. Range Manage. 23(5) :341- 

PHILLIPS, WILLIAM WALLACE. 1967. Prairie soils: 
properties and classification in northwest 
Arkansas. Master's Thesis. University of 
Arkansas. Fayetville. 

TON. 1970. Big bluestem, switchgrass and 
Indian grass. Science and Technology Guide. 
University of Missouri. Columbia Extension 
Division. Columbia. 

SCHRAMM, PETER. 1970. Practical restoration 
method for tall grass prairie. in: Schramm, 
Peter (ed.) 1970. Proceedings of a symposium 
on prairie and prairie restoration. Knox 
College Biological Field Station Publication 
No. 3. Galesburg, Illinois. 

SCHULENBERG, RAY. 1970. Summary of Morton 
Arboretum prairie restoration work, 1963 to 
1968. in: Schramm, Peter (ed.). Proceedings 
of a symposium on prairie and prairie restora- 
tion. Knox College Biol. Field Stat. Special 
Publ. No. 3. 

SNEDECOR, GEORGE W. 1956. Statistical methods 
applied to experiemtns in agriculture and bi- 
ology. Iowa State College Press, Ames, Iowa. 
534 pp. 

STATEN, H. W. 1943. Seeding native grasses. 
Okla. Agr. Exp. Stat. Cir. 108. 

STOIN, HARLAN. 197 5. Personal communication. 

TONEY, TOM. 1975. Personal Communication. 

TURNER, LEWIS M. 1935. Notes on forest types 
of northwest Arkansas. 

Climates of the States, Vol. II. National 
Oceanic and At mospheric Administration Water 
Information Center, Inc., Port Washington, N.Y. 



V. Krupa, R. J. Kohut and J. A. Laurence 

Air pollution is a source- transport-effect 
phenomenon. Knowledge concerning the importance 
of air pollutants and their effects on the human 
environment is continuously increasing across the 
U.S. In the last few years several investigations 
have shown that air pollutants and their potential 
effects are not restricted to the urban areas 
only. A significant amount of data has been 
gathered to demonstrate long distance transport of 
air pollutants and/or their precursors to the 
rural environment. The decline of ponderosa pine 
in the San Bernardino mountains due to the urban 
plume generated in the Los Angeles Basin is an 
excellent example of such a phenomenon (Miller 
and Taylor, 1976) . 

Air pollutants influence our environment 
through their effects on: 1) terrestrial vege- 
tation, 2) aquatic systems, 3) soils and 4) ma- 
terials. Air pollutants occur as gases, aerosols, 
large particulates, and wet fallout (rain and 
snow) . The major sources of national air pollu- 
tion are summarized in Table 1. 

In general, ozone and sulfur dioxide are con- 
sidered to be the most important phy topathogenic 
air pollutants. Plants respond rapidly to these 
and other air pollutants. Based upon their com- 
parative response, plant species are classified 
as sensitive, intermediate and resistant to a 
given air pollutant. A list of some plants known 
to be sensitive to ozone, sulfur dioxide and 
gaseous fluoride is presented in Table 2. Vari- 
ation occurs in plant response to air pollutants 
at the community, generic, species and varietal 
levels. Further, plant age and stage of growth, 
aerial and edaphic environmental factors influence 
the plant response. The research has emphasized 
the effects of air pollutants on economically 
important crop, ornamental and tree species; very 
little is known about the response of herbaceous 
native vegetation to air pollution. 

Air pollutant effects on vegetation may be 
acute, chronic or subtle. Acute effects general- 
ly result from the exposure of plant species to 
high concentrations of air pollutants over few to 
several hours, while chronic effects result from 
exposure to lower air pollutant concentrations 
over days or weeks. While the effects of acute 
and chronic injury are evident in many cases 
through specific symptoms, subtle effects are 
evaluated only as reductions in growth, produc- 
tivity, reproduction and community decline. 

Wester, of the National Park Service, (personal 
communication) made extensive observations of the 
effects of air pollutants on plants in the Capital 
National Parks. However, no detailed studies have 
been conducted on air pollution-vegetation effects 

Paper No. 9792, Scientific Journal Series, 
Minnesota Agricultural Experiemtn Station, St. Paul, 
Minnesota 55108. 
Department of Plant Pathology, University of 

Minnesota, St. Paul 55108. 

with reference to park management. The long-term 
goal of our investigation is to conduct a compara- 
tive investigation of the air quality and vegeta- 
tional effects in a remote and an urban park 
system. The specific objectives are as follows: 

1. to monitor continuously over several 
vegetational growth seasons, ambient 
concentrations of ozone, oxides of 
nitrogen and sulfur dioxide, 

2. to establish permanent vegetational 
study plots and evaluate for symptoms 
of air pollutant-induced injury, 

3. to evaluate selected plant species in 
these study plots for the foliar ac- 
cumulation of total sulfur, fluoride 
and metals, over several growth seasons, 

4. to monitor the changes in soil chemistry 
within the study plots with reference to 
the aforementioned elements, 

5. to evaluate the air pollution response 
of indicator plant species introduced 
into the study area, over several 
growth seasons, and 

6. to evaluate the possible relationships 
between air quality and effects on 
terrestrial vegetation. 


The Voyageurs National Park is relatively new 
and unused and is located in northern Minnesota 
approximately 15 miles east of International 
Falls, Minnesota. This area is detached from 
urban influence and is covered by a hardwood- 
coniferous forest. The understory is composed of 
over 100 different plant species. Before one can 
assess the contribution of man's activities to the 
atmospheric loading of air pollutants in such a 
remote park, it is essential to characterize the 
baseline levels of various air pollutants. In 
this context, ozone and oxides of nitrogen are 
considered to be most important because of the: 
1) remoteness of the Park, 2) generation of ozone 
through natural processes, and 3) possible 
increase in the automobile traffic (a source for 
the ozone precursors - oxides of nitrogen and 
hydrocarbons) . 

During the summer of 1974, ambient ozone con- 
centrations were monitored in the park over a 
period of 676 hours, oxides of nitrogen for 491 
hours and sulfur dioxide for 824 hours (Laurence 
■ t al . , 1975). Throughout the period of obser- 
vation, sulfur dioxide was not detected. Ozone 
concentrations ranged from 0.0075 ppm to 0.08 ppm. 
The latter concentration occurred once during the 
summer for 20 minutes. Mean hourly concentrations 
were lowest from 4 a.m. to 7 a.m. (0.025 ppm) and 
reached a maximum between 2 p.m. and 8 p.m. (0.035 
ppm) (Figure 1) . There were no correlations be- 
tween ozor.e concentrations and frontal systems, 
storm systems or wind direction. During this 
period, the average hourly concentration of total 


nitrogen oxides ranged from 0.32 ppm to 0.397 ppm 
(Figure 1) and the nitrogen dioxide levels varied 
daily from 0.0 ppm to 0.1 ppm. 

It is considered that the 
measured in the Voyageurs Na 
the result of man-related ac 
tance transport. The daily 
indicative of typical photoc 
is suggested that the precur 
tion, both oxides of nitroge 
are produced by the potentia 
activity in the ecosystem 
ozone concentrations are con 
what is generally accepted a 

sampled for foliar chemical analysis of total 
sulfur, fluoride and metals. Soil was also sampled 
for similar chemical analysis. These chemical 
analyses are currently in progress. 

ozone concentrations 
tional Park are not 
tivities or long-dis- 
ozone pattern is 
hemical oxidation. It 
sors for ozone genera- 
n and hydrocarbons, 
lly high biological 
The measured hourly 
sidered to represent 
s background levels. 

During the summers of 1974 and 1975 no symptoms 
of air pollutant-induced vegetational injury were 
observed in the Voyageurs National Park. 


The urban National Park, the Indiana Dunes 
National Lakeshore, is located along Lake Michi- 
gan in northern Indiana. This area presents a 
wide variety of habitats ranging from young un- 
stable dunes and interdunal ponds, to mixed hard- 
wood forests on fully stable ancient dunes. 
During the late 1800 's Henry Cowles conducted his 
classical ecological studies here, which showed 
that plant communities changed over time in pre- 
dictable patterns. This observation led to his 
concept of dynamic succession. This geographic 
area continues to serve an important function in 
research for midwest ecologists, botanists and 
zoologists. The Lakeshore also serves as a pre- 
mier recreation center for the population of the 
Chicago-Gary-Michigan City urban complex. 

It is extremely important to understand the 
effects of the air pollutants in the urban plume 
on the plant communities of the National Lakeshore. 
Air pollutant stress can alter the plant community 
composition by eliminating the sensitive species. 
In addition, air pollutants can also modify the 
rate of plant community succession. 

There are several , diverse air pollutant sources 
in the vicinity of Indiana Dunes National Lake- 
shore. To the west of the Lakeshore, metropolitan 
Chicago serves as a major source for the urban 
plume. The most important phytopathogenic air 
pollutants from this area are considered to be 
ozone and the oxides of nitrogen. The Gary- 
Hammond area west of the Lakeshore is intensively 
industrialized. The myriad industries in this 
area probably contribute a vast variety of emis- 
sions to the atmosphere. Sulfur dioxide, gaseous 
fluorides and various forms of heavy metals are 
considered to be the major air pollutants from 
this are. In addition, several major point sources 
are located in the immediate vicinity of the 
National Lakeshore. These consist of two coal- 
fired power plants and a steel mill. 

During the spring of 1976, a continuous ambient 
air pollution monitoring station was established 
within the National Lakeshore. Ozone, oxides of 
nitrogen and sulfur dioxide were monitored over 
the vegetational growth season (April through 
August) . 

To de .ermine the effects of these pollutants 
on vegetation, 24 field evaluation sites were es- 
tablished in the Dunes area. Vegetation in these 
study plots was evaluated every month during the 
study period for air pollution-induced injury. 
Selected plant species in these plots were also 

Since the plant 
Dunes National Lake 
air pollution stres 
time, it seems poss 
members of the plan 
eliminated. If thi 
of the existing ind 
will be those with 
response to air pol 
tive of the origina 
the evaluation of a 
and their future tr 
pollutant sensitive 
ash and hybrid popl 
study area. These 
several growing sea 

communities in the Indiana 
shore have been subjected to 
s for a substantial period of 
ible that the most sensitive 
t communities may have been 
s process has occurred, many 
ividuals of a plant species 
greater resistance and their 
lutants will not be representa- 
1 plant community. To aid in 
ir pollutant-vegetation effects 
ends, a limited number of air 

plant species such as white 
ar were introduced into the 
plants will be monitored over 
sons for their response to air 

To determine whether the air pollution stress 
has altered the pollution sensitivity of the plant 
in a population of a given species, plants grown 
from seeds collected at the Indiana Dunes National 
Lakeshore are being exposed to ozone and sulfur 
dioxide in controlled laboratory experiments. The 
average levels of injury produced on these plants 
will be compared with the response of plants growr 
from seed collected from areas not subjected to 
the air pollution stress. Air pollutant-response 
studies are also being conducted on a number of 
herbaceous native plant species obtained from out- 
side the Indiana Dunes area. 

During the summer of 1976, ozone injury was ob- 
served on vegetation in 12 of the 24 study plots 
in the Indiana Dunes National Lakeshore area. 
These 12 plots were distributed at random, three 
within the National Lakeshore, one within the ad- 
jacent state park and the remaining eight outside 
the park system. Plant species exhibiting ozone 
injury (Figure 2) included witch hazel, grape, 
strawberry, poison ivy, elm, milkweed, blueberry, 
ash, tulip-poplar, hickory, spicebush, aspen and 
soybean. Sulfur dioxide injury (Figure 2) was ob 
served on four field sites located outside the pa 
system. This injury was observed on blackberry, 
American hazel, elm and nightshade. 

Since the sulfur dioxide injury appeared to be 
rather common at one of the field sites, a series 
of supplementary evaluation sites was establishec 
nearby to determine the spatial extent of the in- 
jury. Species injured included blackberry, ash, 
sweet cicely, boxelder, violet, cherry and bass- 
wood. The sulfur dioxide injury occurred in a b; 
approximately 3.5 miles to 5.0 miles south of an 
industrial complex consisting of a steel mill an- 
a coal-fired power plant. Sufficient air qualit 
and meteorological data are not available at thi 
time to determine the actual source (s) of this 

In general, during the study period elevated 
concentrations of sulfur dioxide were not detect 
in the atmosphere at the air quality monitoring 
station within the National Park. Atmospheric 
concentrations of sulfur dioxide were usually le 
than the expected phytotoxic concentrations. 
Average hourly sulfur dioxide concentrations ex- 
ceeded 0.1 ppm over less than 10 hours during th 
vegetational growth season. It seems likely the 
the location of the air quality monitoring stat: 
was not appropriate to receive the elevated sul- 
fur dioxide concentrations which must have oc- 
curred to produce the vegetational injury at th> 



Major sources of national air pollution. 





Generation of electricity 

Space heating 

Refuse disposal 






"From Middleton (1967) 

TABLE 2. Some plant species known to be sensitive to ozone, 
sulfur dioxide or fluoride air pollutants . 


Sulfur dioxide 




Bent grass 

Brome grass 


Eastern white pine 


Ponderosa pine 

Quaking aspen 

Silver maple 

White ash 

Wild buckwheat 


American elm 

Big-leafed aster 



Curly dock 

Eastern white pine 

Hybrid poplar 



Ponderosa pine 



Blue spruce 



Douglas fir 

Eastern white pine 


Lodgepole pine 

Ponderosa pine 


St. John's-wort 

Western larch 

From Jacobson and Hill (1970) 









FIGURE 1. Ambient 0. 



concentration at Voyageur ' s National Park. 


FIGURE 2. Air pollutant effects on vegetation in the Indi-ia National Lakeshore 
area. (left) Blackberry showing symptoms of S0 2 injury. Note the typical inter- 
veinal necrosis. (right) Dogwood showing upper surface stipple, a symptom of 3 







/• \ 



JUNE M, 1976 /* \ 


/ ''^ I 




/ / x \ 


/ ' v -~- \ 

^ 6 

/ ' ^ \ 



/ ■" \ \ 


a: 5 

1 ' » V 

/ / > A 



/ / » / »\- 


/ I * / * X 


> / JULY ||, 1976 \ ♦ \ \ 
/ i K f \ 



7 1 x ^ ' N 

LU 3 

/ i •*-* \ 


' P * 



/ » 


/ \ 



\ \ 


V^ N * 










1 1 1 1 1 




FIGURE 3. Ambient 0- concentration in Indiana Dunes National Lakeshore. 


TABLE 3. The percentage of plants injured following exposure to a given dosage of ozone. 

Ozone Dosage 

Plant species 







2 hr 

3 hr 

4 hr 

2 hr 

3 hr 

4 hr 


























2 5 













Dwarf columbine 

Giant columbine 



Monkeyf lower 




Polygonus convolvulus 

Cicorium intybus 

Chrysanthemum 1 eucanthemum 

Aquil egia 

Aguil egia 

Chenopodium album 

Lupinus angus ti fol ius 

Mi mul us 

Brassica kaber 

Ambrosia artemisifolia 


Kohut and Krupa (unpublished) 

field sites. It can be concluded that the 
average atmospheric levels of sulfur dioxide must 
have reached at least 0.3 ppm for several hours 
to cause injury on blackberry (a sensitive spe- 
cies) at the field sites. 

Average hourly ozon 
quality monitoring sta 
0.15 ppm. Exposure of 
tions of 0.15 ppm ozon 
generally result in in 
centrations at Indiana 
the diurnal cycling as 
generation (Figure 3) . 
the hourly concentrati 
tion and the pattern d 
urban ozone generation 
oxides of nitrogen wer 
sidered to be phytotox 

e concentrations at the air 
tion were always less than 

vegetation to concentra- 
e for one hour will not 
jury. The hourly ozone con- 
Dunes frequently showed 
sociated with urban ozone 

On certain days, however, 
ons showed little fluctua- 
id not correspond to the 
Concentrations of the 
e below those levels con- 


Some of the herbaceous native plant species 
exposed to ozone under controlled conditions in- 
clude wild buckwheat, chicory, daisy, dwarf colum- 
bine, giant columbine, lambsquarters, lupine, 
monkeyf lower , mustard, ragweed and ribes. The 
average vegetational injury level did not exceed 
10% in any of these exposures. Wild buckwheat, 
chicory, daisy, mustard and ribes were the species 
with the greatest percentage of sensitive plants 
in the population exposed. The symptoms produced 
on the above plants and lambsquarters and monkey- 
flower consisted of a yellow to light tan bifacial 
necrosis. Dwarf columbine exhibited a purple 
stipple. Giant columbine, lupine and ragweed 
were resistant in all of the exposures. Average 
ozone concentrations during the exposure was 0.08 
ppm or 0.15 ppm for 2, 3, or 4 hours (Table 3). 


Two national parks comprising diverse ecosys- 
tems have been examined in the present study on 
air pollution-vegetation effects. The Voyageurs 
National Park is located in northern Minnesota in 
a remote area relatively unaffected by man's ac- 
tivities. Ambient pollutant concentrations in 
that Park were representative of the generally ac- 
cepted background levels. No visible impact of 
air pollutants on the Voyageurs Park vegetation 
was observed. At this time the aforementioned 
studies at the Voyageurs National Park have been 

discontinued. However, as this Park system de- 
velops further, it would be beneficial to re- 
establish the study to evaluate the increased an- 
thropogenic activity in the region in terms of park 
management. Since the vegetation in the Voyageurs 
National Park has not been previously subjected to 
air pollution stress, it can be anticipated that 
the initial effects of increased air pollution on 
the native plant communities could be quite 
noticeable. Although few pollutant sources exist 
in the general area at this time, future mining 
and smelting operations nearby could have a sig- 
nificant impact. 

In contrast, the Indiana Dunes National Lake- 
shore represents an urban ecosystem which has been 
subjected to an air pollution stress for a sub- 
stantial period of time. Our study provided the 
first documented evidence of air pollution injury 
to vegetation in the Indiana Dunes area. Although 
this geographic region has been under air pollu- 
tion stress for some time, observation of the vege- 
tation in the study area indicates that ambient 
pollutant concentrations are sufficiently high to 
injure the vegetation and that the selection 
process by air pollution may be still in progress. 
Additional air quality and meteorological data 
would be beneficial in determining the relation- 
ship between ambient air pollutant concentrations 
and vegetational effects. This information is 
vital for the park management and the implementa- 
tion of emission standards through the state 
regulatory agency. To accurately assess the air 
pollutant effects on the vegetation in the Indiana 
Dunes National Lakeshore, a long term study is 
critical . 

tion of air 
A pictorial 

1975. Ambie 
genie air po 
Park, Minnes 

Special Rep 
16 p. 

MILLER, P. R. a 
of predomina 
ecosystem to 
Preprint Air 


. and HILL, A. C. 1970. Recogni- 
pollution injury to vegetation: 
atlas. Air Poll. Control Assoc. 

. , SCHREIBER, M. C. and KRUPA, S. \ 
nt concentrations of phytopatho- 
llutants in Voyageurs National 
ota. Proc. Amer. Phytopah. Sco. 

1967. The case for clean air. 
Mill and Factory. Conover-Mast . 

nd TAYLOR, O. C. 1976. Response 
nt species in a coniferous forest 

chronic oxidant exposure. 

Poll. Control Assoc. 76-25. 




Walter L. Loope and Neil E. West 


Canyonlands National Park is centered near the 
confluence of the Green and Colorado Rivers in 
southeastern Utah. The Park was established in 
1964 and enlarged to its present 1303 km 2 (337,000 
acres) in 1971. The Park's major scenic resources 
are its dramatic erosional landforms and brightly 
colored rock strata. The geologic section includes 
mostly horizontal Mesozoic sandstones and shales. 

The P 
has been 
one domi 
or absen 
From a c 

ark i 
ry , 1 
t. S 
1 pho 
d, an 

s loc 
965) . 

by r 
een f 
to) , 

d bru 

ated near 
a tectoni 
The Can 
ock. Rai 
ic. Soil 
rom a dis 
age, vast 
shland ar 

the ce 
cally s 

nfall i 

tance ( 
n is ha 

e evide 

nter o 
s land 
s low, 
for ex 
rdly n 
of wo 

f the 
area that 
ne time 
scape is 
and its 
is weak 
ample , a 
odlands , 

Livestock use of the more accessible portions 
of the area were intense in the 60 to 80 years be- 
fore a park was designated and domestic grazing 
was phased out. The area has been a center for 
uranium-vanadium exploration and development. 
Fairly extensive exploration for petroleum has 
also occurred. As development of the master plan 
for the Park began a need for understanding vege- 
tation and its dynamics became apparent. The 
following report highlights some of our findings 
during studies of the flora and vegetation in re- 
lation to environments of Canyonlands National 
Park. Further details can be found in Loope (1977) . 


Reports on vegetation of southeastern Utah are 
few, therefore, our studies had to be done on a 
first approximation level. 

Initially, our questions about the vegetation 
were as follows: 

1. What is the species composition, structure, 
and pattern of the various vegetation types? 

2. What is the location of and total area in 
each vegetation type? 

3. What major environmental factors control 
patterns of vegetation? 

4. How have human activities affected the 

In this landscape dominated by rock, aspects of 
physical environment define several major differ- 
ences in plant habitat. It appeared reasonable 
to use different combinations in these physical 

The encouragement of David May and partial sup- 
port from the Canyonlands Natural History Associa- 
tion is gratefully acknowledged. 

Department of Range Science and the Ecology 

Center, Utah State University, Logan 84322. 

factors to divide the Park into major environmental 
subdivisions for sampling purposes. Substrate 
type, regolith depth, elevation, slope, exposure, 
and position in relation to the water table were 
combined as criteria for identifying landscape 
segments. The resulting subdivisions were: 


Low elevation benches underlain by tight 
clay substrata 

2. Deep eolean sand deposits 

3. Benches underlain by uniformly shallow 

4. Skeletal soils 

5. Benches of alluvial deposits 

Because of the nature of the geologic section, 
some of these occur at one general elevation, 
some at several elevations. 

After environmental subdivisions were es- 
tablished, an effort was made to characterize the 
vegetation of each subdivision through sampling. 
We felt it desirable to include the entire park in 
the analysis. Limited time and budgets, large 
area considered, dissected topography, and diffi- 
cult access forced some sacrifices in objectify- 
ing the placement of sampling sites. Sites with- 
in each environmental subdivision were selected 
as "typical." Sampled sites were scattered 
throughout the Park. 

After selection of each sampling point, a 100 
meter steel tape was extended from the point in 
the direction of the most visually homogeneous 
vegetation. An estimate of vegetal cover was 
taken as the percent intercepted cover of plant 
canopies along the 100 meter tape. Twelve small 
plots were positioned at random intervals along 
the steel tape. Frequency was recorded as the 
percent of the small plots where a given species 
was present. Frequency plots of 1 m2 size were 
used for herbaceous vegetation; 4 m2 plots were 
used for non-herbaceous plants. Density was es- 
timated by a count of individuals in a 5 m x 50 m 
rectangular macroplot positioned along the tape 
with one corner at the sampling point. In a sub- 
sample of the sites within each environmental 
subdivision, measurements of soil texture, pH and 
total electrical conductivity were taken. Slope, 
exposure, geologic formation, regolith depth, 
elevation and position in relation to the water 
table were noted for each site. Constancy values 
for all species were calculated as the percent of 
transects within each environmental subdivision 
in which the species occurs. One hundred fifty- 
seven transects were read. 


Results of sampling are shown in condensed form 
in Table 1. "Importance index" (constancy x 
frequency) (Curtis, 1959) was calculated for all 
species. Species with an importance index greater 
than 1000 are included in the table. It is clear 
from Table 1 that distinct groups of species 


typify each environmental subdivision. 

The environmental subdivision defined by out- 
croppings of shale with high clay content (tight 
soil benches) supports a sparse shrubland dominated 
by several species of atriplex. The community 
composition varies from dominance by A. garrettii 
on most sampled areas to a. cuneata and A. 
conferti folia at other sites within this environ- 
mental subdivision. 

On deep eolean sand deposits, grasslands occur 
almost exclusively. Hilaria jamesii, Stipa 
comata and Oryzopsis hymenoides are grasses found 
to have highest importance values. The shrubs, 
Ephedra viridis and Ceratoides lanata are common 
associates in some areas. Deep eolean deposits 
that incurred historically heavy livestock graz- 
ing use were found to support a modified grassland 
vegetation. On these sites, the grazing-resistant 
grasses, Hilaria jamesii, Sporobolus cryptandrus 
and Aristida fendleriana are more important. 
Gutierrezia spp. are also more common in these areas 
than those grasslands that were lightly grazed by 
livestock or escaped livestock use altogether. 

Coleogyne ramosi ssima-dominated brushland 
occupies most sites where regolith is uniformly 
shallow. The shrubs, Ephedra torreyana and 
Atriplex conferti folia , and the grass, Hilaria 
jamesii, are the most important subordinates in 
this community. 

Jointed, exposed bedrock areas with skeletal 
soils support pinyon- juniper woodlands with 
various upland shrubs in the understory. Vegeta- 
tion is generally confined to cracks in bedrock. 
Coleogyne ramosi ssima and Artemisia bigelovi i 
occur frequently with Pinvs edulis and Juniperus 
os teosperma . 

Deep alluvial deposits with access to capillary 
water are characterized by brushlands dominated 
by Artemisia tr identata and Atriplex canescens . 

Two distinctive vegetation types were not 
sampled like the others because of their very 
limited area. Species lists and collections of 
vouchers were made during visits to these habi- 
tats, however. These minor vegetation types are 
the dense riparian thickets along the Green and 
Colorado Rivers and the "hanging gardens" that 
occupy topographically favored alcoves and over- 
hangs supplied with moisture from seeps. The 
dense thickets of vegetation along the rivers de- 
pend on free water at their roots and are composed 
chiefly of Tamarix pentandra, Pluchea sericea, 
Baccharis emoryi and several species of Salix. 
"Hanging gardens" are characterized by diverse 
groupings of mesic species such as Aquilegia 
micrantha, Primula specuicola, Mimulus eastwoodiae 
and Smilacina stellata . Betula Occident al i s , 
Prunus virginiana and Ostrya knowltoni are mesic 
trees that often occur in these situations. 

Dark-colored cryptogamic soil crusts, com- 
posed of lichens, mosses, fungi, and algae are 
widespread throughout Canyonlands National Park. 
Their distribution is partially a function of 
livestock use history. Heavily used areas gen- 
erally have low cover values of cryptogamic soil 
crust. In some areas substrate appears to be im- 
portant in controlling their distribution. Sandy 
soils have more of these crusts than clayish or 
silty soi^ surfaces. 

After major vegetation units were defined, 
type signatures were identified on aerial photos 
and a vegetation map of the Park was made. A 15 
minute topographic map was used as the base. This 

map and a full floristic list for the Park may be 
found in Loope (1977) . 

The distinctness of vegetation units suggests 
sharp environmental gradients. In areas were 
evapotranspiration greatly exceeds precipitation, 
soil moisture will generally be a very important 
environmental factor. Effective soil moisture ap- 
pears to be a critical controlling factor for 
vegetation patterns in Canyonlands. 

In many areas of the western United States soil 
moisture is chiefly a function of elevation, slope 
and exposure. The most mesic sites occur at high 
elevation on north-facing slopes with a gradient of 
increasing dryness toward lower elevation, south- 
facing slopes. These gradients have a much smaller 
influence in Canyonlands National Park. Effective 
soil moisture in Canyonlands appears to be pri- 
marily controlled by regolith depth and position 
in relation to the water table. 

Effective soil moisture can be related to 
regolith, relationships in the following ways (Fig- 
ure 1) . When disintegrated bedrock is uniformly 
shallow (ca. 25 cm) , soil moisture is concentrated 
on top of impermeable bedrock at a shallow depth. 
This perching effect allows time for gradual up- 
take of moisture by plant roots. Blackbrush domi- 
nates the vegetation in this situation. Hunt (1966) 
similarly observed impermeable layers causing mois- 
ture concentration and concomitant distinctions in 
the desert vegetation patterns of Death Valley, 
California. When regolith is deeper than 50 cm, 
this perching effect is lost and blackbrush gives 
way to grassland vegetation. Grass root systems 
are apparently better adapted to take up rapidly 
percolating soil moisture, particularly during 
growing season rainy periods. In areas where rego- 
lith is restricted to bedrock fissures, pinyon- 
juniper/shrub vegetation occurs. These sites get 
a "moisture subsidy" of runoff from areas between 
cracks and thus support the relatively mesic 
pinyon-juniper vegetation. Tight clay substrate 
restricts movement of moisture to plant roots and 
thus limits rooting to a shallow soil profile. 
Outcroppings of this type of substrate controls 
the occurrence of Atriplex-dominated vegetation. 

Access versus non-access to free or capillary 
water controls the occurrence of species in 
riparian and hanging garden habitats. 

The pattern of regolith-controlled soil moisture 
and associated vegetation types remains distinc- 
tive and repeats itself at various elevations with- 
in the Park. Figure 2 illustrates this recurring 
pattern. In Canyonlands, the influence of eleva- 
tion on effective soil moisture is apparently over- 
shadowed by the influence of regolith depth. 

Fire does no 
maintenance of 
of Canyonlands. 
the landscape, 
often separated 
Blackbrush, a s 
(Bowns and West 
centage of the 
to the pinyon-j 
rarely enough f 
scape . 

t appear to play 

the brushlands or 

Topographic bre 

Isolated pockets 

by rock walls or 

pecies that is in 

, 1975) , dominate 

Park. Neither is 

uniper woodlands 

uel to carry fire 

role in the 

eolean grasslands 
aks highly dissect 

of vegetation are 

steep canyons, 
tolerant of fires 
s the largest per- 

fire important 
since there is 

across the land- 

Soils in Canyonlands display very weak develop- 
ment (Wilson et al., 1975). Annual rainfall is 
very low (averaging approximately 210 mm) . Most 
substrate is fine quartz sand or rocks that weathe C 
to fine quartz sand. The distinctive topography 
of the Park has been formed by rapid stripping of 
the rock mantle. Material is eroded and 


Grassland , Pinyon-Juniper t Blackbrush | Atriplex 

Deep eolean Skeletal 

sand soil 


sh.ll low 

Tight shale 

FIGURE 1. Regolithic control of soil moisture and vegetation types. 



Vegetation Types 


P/.T Shruh m.- u -.kbrush r Grasslan d 

/' "■ ' v ' (Kayenta) 

Skeletal Uniformly Shallow Deep Lolean 
Soil Regolith Sand 

I W iiiL-it e I 

Grassland P/J Shrub Blackbrush Atriplex 



Deep Eolean Skeletal Uniformly Shallow Tight 
Sand Soil Regolith Soil 

FIGURE 2. Vegetation types controlled by regolith depth recurring at various elevations. 


TABLE 1. Percent frequency (F) , density (d, common plants/m ) , and percent cover (C) values for species 
with constancy X frequency index greater than 1000. Environmental subdivisions are arranged to show an 
effective soil moisture gradient from xeric (left) to mesic (right). The symbol, t, indicates less than 
1% cover (trace) . 

Palant name* 

Environment Subdivision 

Low elevation 
tight soil 
F D C 

Eolean sand 
(heavy use) 
F D C 

Eolean sand 
(light use) 
F D C 

1 4 
6 6 

2 r > 

Artemisia tridentata — big sagebrush 

Atriplex canescens--four wing saltbush 

Opuntia spp. — prickly pear 

Chrysothamnus nauseosus — rubber rabbitbrush 

Sarcobatus vermicula tus--black greasewood 

Lepidium fremontii — desert pepperweed 

Bromus tec tor um — cheatgrass brome 

Sporobolus contractus— spike dropseed 

Descurainia sophia — flixweed tansymustard 

Aster bigelovii — Bigelow aster 

Lappula redowskii — bluebur stickseed 

Juniperus osteosperma — Utah juniper 

Pinus edulis — pinyon pine 

Artemisia bi gelovi i--Bigelow sagebrush 

Hymenoxys acaulis — stemless actinea 

Mac hae rant her a grindeloides — goldenweed 

Coleogyne ramosissima — blackbrush 

Ephedra torreyana--Torrey mormon tea 

Atriplex conferti folia — shadescale saltbush 

Hilaria jamesi i--galleta 

Ephedra viridis — green mormon tea 

Ceratoides 1 ana ta (Pursh) J.T.Howell — winterfat 

Stipa comata--needle and thread 

Oryzopsis hymenoides — Indian ricegrass 

BouteJoua gracil is--blue gramma 

Blantago patagonica — wooiy plantain 

Eriogonum cernuum — nodding wild buckwheat 

Gilia inconspicua — shy gilia 

Gutlerrezia spp. --snakeweed 

Cryptantha confertiflora — cryptantha 

Shaeralcea coccinea--scarlet globemallow 

Hymenopappus f i 1 i f ol i urn — f ineleaf hymenopappus 

Sporobolus cryptandrus — sand dropseed 

Aristida fendleriana--Fendler threeawn 

Vulpia octoflora (Walt.) Rydberg--sixweeks fescue 

Malacothrix sonchoides--desert dandelion 

Oenothera caespitosa — tufted evening primrose 

Cymopterus fendleri--Fendler rpring parsley 

Atriplex garrettii — Garrett saltbush 

Eriogonum inf latum — desert trumpet 

Gai llardia pinnatafi da --blanket flower 

Salsola kali — Russian thistle 

Phacel ia crenu 1 a ta--scorpionweed 

*Nomenclature in the table and text follows Holmgren and Reveal (1966) except where authorities 
are given. Common names follow Beetle (1970). 




r > 




2 1 



6 1 


J 9 




r .7 







i r > 





4 I 



3 1 














4 6 


4 4 


2 6 




r ,4 



J 4 










transported out of the area fa 
modified (Hunt, 1974). It is 
surface soil texture is more a 
strate than of soil developmen 
was undetected in the profiles 
authors. Neither were differe 
total conductivity or surface 
ments, except that textural an 
from tight soil benches showed 
silt and clay than other sampl 
horizons (caliche), common in 
Desert, were not found in soil 
lands . 

ster than it can be 
probably true that 

function of sub- 
t. Horizonation 

observed by 
nces obvious in pH, 
texture measure- 
alysis of samples 

a higher percentage 
es . Petrocalcic 
the Great Basin 

profiles of Canyon- 

tal cover in some areas with historically heavy 
livestock use are made up of cheatgrass (Bromus 
tectorum) , sixweeks grass Vulpia octoflora), 
Russian thistle (Salsola kali) and snakeweed 
Gutierrezia spp.) Grassland modifications are 
probably the result of differences in accessibilit 
to early cattlemen. Grasslands of the southeaster 
portion of Canyonlands National Park were, by far, 
the most accessible. Elsewhere, grasslands are 
isolated by steep cliffs or broken topography. 

Grasslands have undergone more change due to 
human activities than other vegetation types with- 
in the Park. Evidence of grassland disturbance 
by domestic livestock use is greatest in the south- 
eastern section (Needles District) of Canyonlands 
National Park. Other areas display much less 
modification. High percentages of the total vege- 

Disturbance due 
petroleum explorati 
tion types. Total 
ly small, but it is 
Secondary plant sue 
extremely slow. Ho 
blackbrush vegetati 
essentially bare 20 
areas where blackbr 
invaded by exotics 

to uranium-vanadium and 

on has occurred in all vegeta- 

area of this damage is relative 

spread throughout the Park, 
cession on these sites has beer 
st old access roads through 
on on shallow regolith remain 

years after abandonment. Some 
ush was destroyed have been 
such as Halogeton glomeratus 











F D 












1 4 








2 8 









2 2 















76 76 





25 10 


23 11 









taken at each site, and soil and geologic parame- 
ters were noted. Abandoned roads and drill sites 
were examined for evidence of secondary succession. 
Two hundred thirty vascular plant species were 
collected in the Park. A vegetation map of the 
Park was constructed using aerial photographs. 

The results of this study support the following 
conclusions : 

1. There are five major vegetation types 
within the Park. Each type appears 
dependent on specific environmental 
conditions . 

2. Soil moisture is the key factor influencing 
vegetation distribution in Canyonlands. 

3. Major control over soil moisture is exerted 
by regolith depth and water table relation- 
ships . 

4. Secondary control over soil moisture (and 
thus vegetation patterns) is exerted by 
geologic substrate, elevation, slope and 

5. Secondary plant succession is very slow in 
all vegetation types. In shallow regolith 
situations, secondary succession, in the 
sense of site preparation by serai plants, 
may not occur at all. 

6. Livestock use has caused measurable altera- 
tions of grasslands on eolean deposits in 
Needles District. 

7. Fire has not been a pervasive or important 
factor in the maintenance of any vegetation 
type in Canyonlands. 

8. The physical environment is dominant in 
determining vegetation structure and dis- 
tribution in Canyonlands National Park. 



and Salsola kali or aggressive native species such 
as Gutierrezia spp. and Opuntia spp. Secondary 
succession appears to be more related to the 
physical recovery of the site than with any modi- 
fication by vegetation. Road cuts through deep 
eolean material appear to recover faster than cuts 
through skeletal soil or shallow regolith. Re- 
vegetation efforts, where attempted, have met with 
very limited success. 


The purpose of this study was to provide a 
description of the flora and vegetation of Canyon- 
lands National Park and to relate vegetation types 
to major environmental variables within the Park. 
One hundred fifty-seven transects were located 
throughout the Park to represent major environ- 
mental situations. Areas that underwent his- 
torically heavy grazing were included. Measure- 
ments of species frequency, density and cover were 

BEETLE, A. A. 1970. Recommended plant names. 
Wyo. Agr. Expt. Sta. Res. J. 31. 124 p. 

BOWNS, J. E., and N. E. WEST. 1976. Blackbrush 
(Coleogyne ramosissima ) on southwestern Utah 
rangelands. Utah Agr. Exp. Sta. Res. Rpt. 27. 
30 p. 

CURTIS, J. T. 1959. The vegetation of Wisconsin. 
Univ. Wise. Press, Madison. 657 p. 

HOLMGREN, A. H., and J. L. REVEAL. 1966. Check- 
list of the vascular plants of the inter- 
mountain region. U.S.D.A. For. Serv. Res. Pap. 
INT-32. 160 p. 

Plant ecology of Death Valley, 
. Geological Survey Prof. 

HUNT, C. B. 1966 
California. U 
Paper 509. 68 

HUNT, C. B. 1974 
and Canada. W 



Natural Regions of the V.i 
H. Greeman and Co., San 
725 p. 
1977. Vegetation-environmental 

lationships in Canyonlands National Park, 

Utah. Ph.D. diss., Utah State Univ., 

Logan. 138 p. 
THORNBURG, W. D. 1965. Regional geomorphology 

of the United States. John Wiley and Sons, 

New York. 505 p. 

SOUTHARD, and A. J. ERICKSON. 1975. Soils of 

Utah. Utah Agr. Exp. Sta. Bull. 492. 94 p. 



Ingrid Olmsted 
College of Forest Resources 
University of Washington 
Seattle 98195 

Reprints of this paper are available from Dr. Olmsted. 


William H. Thomas 

Institute of Marine Resources 

Scripps Institution of Oceanography 

University of California, San Diego 

La Jolla 92037 

This paper is published in Journal of Phycology 
8(l):l-9, 1972 



. 1 

W.H. Moir, F.D. Hobson, M. Hemstrom, and J.F. Franklin 


We began studies of forest ecosystems at Mount 
Rainier National Park (MRNP) in the summer of 1975. 
Our objectives were to idenitfy and describe the 
different forest types, characterize their envi- 
ronmental features and patterns of occurrence, 
within the Park, and determine causes and rates of 
succession. The initial forest classification 
presented here is based on data from 242 plots dis- 
tributed over all the major drainage systems. 
During our second summer (19 76) we sampled an addi- 
tional 158 plots, mostly younger forests. This 
paper reports the major forest types resolved from 
these data, some generalized patterns of occurrence 
of these types in selected landscapes at MRNP, the 
major soil types, and important factors initiating 
succession. We also give forest age spectra for 
three drainages and compositional chronosequences 
in three of the forest types. 


Most data were obtained by the 
method for forest classification { 
1970) . The usual procedure for th 
would be to travel a slope, trail, 
or whatever path through a foreste 
establish a string of sampled plot 
both the typical and changing fore 
along that path. In 1975 we sough 
late serai to climax condition, i 
relatively free of recent disturba 
we focused mostly upon forests of 
serai condition. 

Franklin et al. 
e field crew 

road segment, 
d landscape and 
s that revealed 
st patterns 
t stands from 
e. , old-growth 
nces; in 1976 
early to middle 

Circular plots mostly 500 or 1,000 m (depend- 
ing on mature tree density) were our sample units. 
Understory vascular plant cover was recorded as 
described by Franklin et al. (1970) and Dyrness 
et al. (1974). All tree stems exceeding 1.4 m in 
height were tallied by species and diameter. Tree 
regeneration (1.4 m in height) was subsampled in 
a 50 m circular plot either at the center of the 
larger plot (1975) or at four 12.5 m 2 circular 
plots, one in each quadrant of the larger plot. 
We also made field description of soils, and re- 
corded elevation, exposure, slope, degree of 
canopy closure, landform, position in the land- 
scape (lower, middle, upper slope, ridge, draw, 
bench) and location on topographic map. Supple- 
mental notes included disturbance evidence (fire, 
biotic, wind-throw, etc.) and inclusions or 
mosaics of other forest types. 

The classification was obtained by subjective- 
ly sorting releve' or stand table data from late 
serai to climax plots into subsets based upon 
general similarities in dominance within each of 
the tree shrub, and herb strata. The technique is 
described by Dyrness et al. (1974) and Shimwell 
(1971, p. 188 ff ) . 

Stand ages were measured from increment cores 
of dominant specimens of Pseudotsuga menziesii, 
Abies procera, or other trees judged to have 
been among the first wave of regeneration after 

United States Forest Service, Forest Sciences 
Laboratory, Corvallis, Oregon. 

disturbance. We sought the oldest speciments of a 
serai cohort to set a time limit on whatever dis- 
turbance event gave ooportunity to establish that 
cohort. Cored trees were aged by counting rings 
in the field and adding to that age corrections for 
center and age to core height. For large speci- 
mens these corrections were sometimes a source of 
considerable sample error. 


We can recognize four broad elevational zones 
of forests at MRNP. At highest elevations (gen- 
erally over 1,620 m) are subalpine parklands where 
mosaics of three copses and meadow communities oc- 
cur. These communities and environments have been 
previously described (Franklin and Mirchell 1967, 
Franklin et al. 1971) and are not the subject of 
this study. Below about 1,620 m elevation are 
mostly closed canopy forests. The Tsuga merten- 
siana zone occurs at the hightest elevations with 
Ahies lasiocarpa, Tsuga mertens iana , Chamaecyparis 
nootkatensis , and Abies amabilis as characteristic 
trees. At mid-elevations Abies procera, Pseudot- 
suga menzies ii , Tsuga heterophyl la , and Thuja 
plica ta are important canopy dominants. These 
forests are within the Abies amabilia zone gen- 
erally between 730 and 1,475 m elevation at MRNP. 
We have provisionally recognized 3 high-elevation 
and 9 mid-elevation forest types at MRNP (Table 1) . 
Below about 730 m elevation are four forest 
types of the Tsuga heterophylla zone (Franklin and 
Dyrness 1973) . Such forests are more extensive 
outside the Park, but reach their upper elevation- 
al limits in valleys and south-facing slopes near 
Park boundaries. Pseudotsuga mensiesii , Tsuga 
heterophylla , Thuja plicata , Abies grandis , and 
Picea sitchensis are trees of these lowest eleva- 
tions . 

The forests of MRNP can also be divided into 
mature and immature types. The classification of 
Table 1 is based upon mature forests that are 
generally about 250 years or older and well 
stocked. Immature stands are either well below 
stocking potential or less than about 250 years 
old. Serai forest types (or community types) in- 
clude Abies lasiocarpa forests at high elevations, 
Pseudotsuga menzies i i / Pter id i urn aquilnum communi- 
ty types at intermediate elevations, and red alder 
{Alnus rubra) communities of lower slopes and val- 
ley bottoms. 


Moisture and temperature gradients are major 
environmental complexes affecting forest patterns. 
For each drainage system at MRNP these complexes 
can be delimited by elevation, relief, and posi- 
tion in the landscape. Thus each forest type em- 
braces a relatively narrow environmental span 
illustrated by the profiles of Figures 1-3. 

white River. Our profile (Fig. 1) extends 
from Clover Lake to Sunrise road in the White 
River Valley, an elevational range of 740 m 
(2,430 ft). The road to Sunrise Visitor Center 
crosses the entirety of this sequence of forest 
types. Subalpine parkland is common along 


TABLE 1. Major forest types at Mount Rainier. 

Forest Type 

Dominant Trees- 
Over story Regeneration 

Important Under6tory Species 

I. LOW ELEVATION (< 730 m) 

Western hemlock/Salal 
Western hemlock/Vanillaleaf 
Western hemlock/Swordfern 
Western hemlock/Devil's club 






Gaultheria shallon , Vaccinium parvifolium , Berberis nervosa 

Acer circinatum , Achlys triphylla , Viola sempervirens 

Acer circinatum , Polystichum munitum , Berberis nervo sa 

Oplopanax horridum , Athyrium filix-femina . Gymnocarpium dryopterie 

W. hemlock-Silver Fir/Oregongrape PSME, TSHE 

W. hemlock-Silver Fir /Swordfern- 
Deer fern 

Silver Fir/Salal 

Silver Fir/Alaska huckleberry 

Silver Fir/Alaska huckleberry- 
Trailing raspberry 

Stiver Fir/Foam flower 

Silver Fir/Devil's club 

Silver Fir/Devil's club/Trefoil 
foam flower 

Silver Fir/ Hue kleberry/ Bear grass 

Silver Fir-Yellow cedar/ 
Oval-leaf huckleberry 

HIGH ELEVATION (> 1300 m) 

Silver Fir / Hue kleber ry /Herb 

Silver Fir/Rusty leaf 

TSHE, ABAM Acer circinatum , Berberis nervosa , Achlys triphylla , Vaccinium 
PSME, TSHE, THPL TSHE, ABAM Polyetichum munitum , Blechnum spicant , Berberis nervosa 

PSME, TSHE, THPL TSHE, ABAM Gaultheria shallon , Xerophyllum tenax , Berberis nervosa 
TSHE, ABAM ABAM Vaccinium alaskaense , Linnaea borealis , Rubus lasiococcus 

TSHE, ABAM ABAM Vaccinium alaskaense , Vaccinium o valifolium , Rubus pedatus 


Tiarella unifoliata, Achlys triphylla Streptopus roseus , Clintonia uniflora 
Oplopanax horridum , Gymnocarpium dr yopter is, Tiarella unifoliata 
Oplopanax horridum , Vaccinium spp . , Gymnocarpium , R ubus pedatu s, 
Tiarella trifoliata 

TSHE, PSME, ABAM ABAM, TSHE Vaccinium membran aceum , Xerophyllum tenax 

TSHE, CHNO ABAM Vaccinium ovalifolium , Tiarella unifoliata , Gymnocarpium dryopteris 




Vaccinium membra naceum , Erythromum montanum , Rubus pedatus 

Rhododendron albiflorum , Menziesia ferruginea , Vaccinium 

1/ PSME - Pseudotsuga menzlesll , TSHE - Tguga heterophylla . TSHE - T. mertenslana , ABAM - Abies amabtlls , ABPR - A. procera , CHNO « Chamaecyp arls 
nootkatensls, THPL • Thuja pllcata 

Sunrise Ridge and in the upper Sunrise Creek basin 
along the trail to Clover Lake (elevations gen- 
erally over 1,620 m) . A small finger of Silver 
fir/Rusty leaf forest type occurs along Sunrise 
Creek below Clover Lake. Elsewhere on steep 
slopes between 1,500 to 1,620 m are closed sub- 
alpine fir forests, as, for example, along Sunrise 
Road. The Silver fir/Foam flower forest type cen- 
tered around 1,500 m elevation. The boundary is 
visually apparent because subalpine fir (dark 
greenish canopies) suddenly yield dominance to 
noble fir (bluish canopy). At 1,280 m is a nar- 
row, wet bench on which occurs the Silver fir- 
Alaska yellow cedar /Oval-leaf huckleberry type. 
Below this the Western hemlock-Silver fir/Oregon- 
grape forest extends along middle and lower 
slopes. The lower slopes and valley bottoms along 
the White River have forests of Silver fir/Alaska 
huckleberry type, with Silver fir/Devil's club 
type in seep and smaller drainage areas. Occasion- 
al Engelmann spruce and Alaska yellow cedar along 
the White River may indicate cold airflow and 
other environmental features of this major valley 
drainage . 

Ohanapecosh River. The southeastern sector of 
Mount Rainier is the driest of the Park's drain- 
ages — the rain shadow of Pacific westerlies 
swirling around the volcanic cone. Our profile 
(Fig. 2) extends from Cowlitz Divide, across the 
steep, narrow Ohanapecosh Valley to the southeast 
boundary of the Park on the ridge of the Ohanape- 
cosh rock formation. The Steven's Canyon road 
crosses the east-facing slope of this profile and 
offers a bird's eye view of forests across the 
valley. The ridge forming the Park's east boun- 

dary has forests of the Sliver fir/Foam flower and 
Silver fir/Rusty leaf types. Below 1,300 m and 
extending all the way to the valley floor are 
forests dominated almost entirely by Douglas fir 
and western hemlock along west-facing slopes. At 
higher elevations are forests of Western hemlock- 
Silver f ir/Oregongrape type; lower slopes are 
Western hemlock/Salal type. The Ohanapecosh 
River valley contains forest mosaics of Western 
hemlock/Salal, Western hemlock/Vanilla-leaf , and 
Western Hemlock/Devil's club. The very steep can- 
yon sideslopes leading up to Cowlitz Divide gen- 
erally mirror the zonation pattern of the opposite 
slope. Cliffs and rocky talus, however, 
frequently interrupt the forests. Shallow, cobbly 
soil between the cliffs and ledges may exhibit 
xeric variations of Western hemlock/Salal and 
Western hemlock-Silver f ir/Oregongrape forests. 
But Silver fir/Foam flower forest type can be found 
on more stable slopes with deep tephra soils (Fig. 
2 just below the road) . The ridges and upper 
slopes of Cowlitz Divide have forests of Silver 
f ir/Huckleberry/Beargrass and Western hemlock- 
Silver f ir/Oregongrape with the former on more 
exposed sites. 

Mowich River. This westerly trending drainage 
system is (together with the Carbon River drain- 
age) the wettest of the Park's watersheds. Fig- 
ure 3, Paul Peak to the Golden Lakes upland, 
shows great variation of forest pattern from south 
facing to north-facing slopes. Forest sequences 
along slopes of Paul Peak are not unlike the 
west-facing slopes of the Ohanapecosh River (Fig. 
2). The lower slopes along Mowich River, however 
contain stands of the Western hemlock/Swordfern 








- ~ \^$|J££$/ \ 




l\ SE _ 


Y Abies amobilis 
A bias procsro 
Pseudotsugo memiesii 

VVjQ Berbens nervosa — 


Y$&1 Vaccinium alaskaense 


$ Abies fosiocarpo 
__ A Picea engetmannii 

\ST3ji, / Oplopanax horndum 
\ll / SUNRISE 
\M R0AD_^ 


— XL Chamaecyparis nootkatensis 

/Wu 1 (i$m)~ 

I Tsuga heterophylla 



- 1,800 


1,600 * 



1,400 2 






- 1,000 

FIGURE 1. Generalized vegetation cross-section across Sunrise Ridge from 
Clover Lake to the White River Valley. 





"■ 3.600 

i= 3.200 







- J 

(>• Abies a ma bilts k!$/^ 

*/) Abies procera *%*/?+■ TRAIL 

\X (CLIFFS) ~~W/ 


vtL /Road wrr 


\\A Tsuga heterophylla ^'iAlJv' 



WA ROAO iAlJ/^ . . a . 

Wr\ \ A*kS^ Berbens nervosa 
ysjAj \ aAt3®^ l=* Vaccinium alaskaense 
tvLVa ^WtXc^^ FTTTF1 Xerophyflum tenax 
v^jIt itn^^fc^^ UTTTTTTi Gaulthena shollon — 
^3|}Ljtjgii£^ J ^ l?'b1 Ach/ys triphylla 
^ES***^ PW< Oplopanax horndum 

■■ Rhododendron atbiflorum - 


1,200 <* 

- 800 

- 600 

FIGURE 2. Generalized vegetation cross-section of the Ohanapecosh Valley 
near the ranger station and campground. 




w 4,000 



w 3.200 




A bus lasiocarpa 


i i Berbens nervosa 

Vaccinium aloskoense 
FTTH Xerophyllum tenax 
irriffl Gaulthena shallon 
m Polystichum munitum 
Oplopanax horndum 
Rhododendron olblflorum 


1.200 2 



FIGURE 3. Generalized vegetation cross-suction of the Mowich River Valley. 


type. The Mowich River bottomlands are mantled 
primarily by forests of Silver fir/Devil's club/ 
Trefoil foamf lower. Colluvial lower slopes and 
draws with northerly aspect feature the Silver 
f ir/Swordfern-Deerfern forest type. Around 1,100 m 
elevation the trail to Golden Lakes passes through 
stands dominated by western hemlock and silver 
fir, but with a very depauperate, sterile ap- 
pearing understory. Infrequent deerfern or huckle- 
berry give only faint suggestion of the Silver 
f ir/Swordfern-Deerfern or Silver fir/Alaska 
huckleberry forest types. The latter is better 
defined as understory flora becomes more diverse 
and abundant upslope (ca 1,200 m elevation). The 
uppermost north-facing slopes are forested with 
the Silver fir/Rusty leaf type. The ridgetop and 
upperslopes of the Golden Lakes basin area con- 
tain open, fire-derived forests of the Silver fir/ 
Huckleberry/Beargrass type with occasional 
subalpine fir among early successional trees. 


The forest soils at HRNP are extremely variable 
and grade from one kind of soil to another. Ma- 
jor soil-forming activities are accumulation of 
forest floor organic matter and development of 

iron pans. Podzolization may sometimes be in- 
dicated by very weak to moderately-developed 
spodic horizons. Four basic parent materials 
(Hobson 1976) are pyroclastic deposits (tephra) , 
mudflow, colluvium, and alluvium. Tephras, 
alluvial, laharic, and glacial materials have 
periodically been deposited during the history of 
forest vegetation at MRNP (Mullineaux 1974, 
Crandell 1971, 1967), and most parent materials 
of the rooting zone are of Holocene age. Differ- 
ences in soil mineralogy and depths of various 
parent materials within the profile generally 
seem to have little influence upon the composi- 
tion of forest vegetation. But soils considered 
in the broader context of landform, internal 
drainage, and position in the landscape (Table 
2) can affect vegetation. For example, the 
colluvial soils of steep midslopes may be physio- 
logically drier than deeper, fine-textured tephra 
soils of midslope benches. 


Fire is the most extensive forest disturbance 
and its evidence is found in the majority of our 
sample plots. Fires appear to have influenced 
the composition of wetter valley forests as well 

TABLE 2. Modal soil profiles of forests at Mount Rainier National Park (after Hobson 1976) 
1. Tephra-Slope- No Iron Pan 

01 6-4 cm 

02 6-0 cm 
A21 0-3 cm 





3-9 cm 

9-25 cm 

25-32 cm 

32-68 cm 


Duff; abundant medium and fine roots; smooth, abrupt boundary to 
Dark gray (moist) fine sand (post-W tephra); loose; fine and medium 
roots abundant; smooth, abrupt boundary to 

Mixed brown and gray (dry) coarse sand (tephra W); loose grained; 
fine and medium roots common; occasional subangular cobbles; wavy, 
clear boundary to 

Brown (moist) loamy sand (tephra C) with fine distinct areas of dark 
gray fine sand; massive breaking to loose weak, moderate subangular 
blocks, very friable; few roots; 5 percent lapilli, occasional subangular 
cobble; smooth clear boundary to 

Dark gray sand (moist) with common fine distince brown mottles; mas- 
sive breaking to loose and very weak subangular block, very friable; 
few roots; 10 percent lapilli; smooth abrupt boundary to 
Mixed brown, dark gray, and yellowish brown very coarse sand (tephra 
Y); loose; very few roots to none; occasional cobble 

2. Tephra- Bench-Strong Iron Pan 

01 6-4 cm 

02 4-0 cm 
A2 0-7 cm 

Bl 7-18 cm 

B2ir 18-36 cm 

3. Colluvial 


9-7 cm 


7-0 cm 


0-2 cm 


2- 5 cm 


5-60 cm 

4. All 



3-2 cm 


2-0 cm 


0-3 cm 


3-14 cm 

14-63 cm 


Duff; saturated; abundant roots 

Gray (wet) very coarse sand (tephra Y); loose; roots abundant, more 

so at 02/A2 boundary; smooth clear boundary to 

Mixed light brownish gray and dark yellow brown (wet) very coarse 

sand; loose grain; roots common; smooth gradual boundary to 

Dark brown (wet) very coarse sand; massive breaks to structureless, 

friable becoming firm with depth; very few to no roots. 


Duff; roots common 

Dark gray (moist) fine sand (post-W tephra); loose; roots common; 

smooth abrupt boundary to 

Mixed gray and yellowish brown (moist) coarse sand (tephra W); single 

grain; roots common; irregular abrupt boundary to 

Yellowish brown (moist) gravelly sand; 25 percent lapilli and angular 

gravels increasing with depth; loose; roots abundant to 40 cm grade to few. 


Duff; fine roots abundant 

Very dark grayish brown (moist) fine sand; massive breaks to moderate 

medium crumb, friable; fine roots common; smooth clear boundary to 

Dark yellowish brown (moist) fine sand; massive breaks to moderate 

coarse subangular block, very friable; fine roots common; smooth clear 

boundary to 

Dark gray (moist) sand; massive breaks to single grain; few fine and 

medium roots. 


as those of drier ridges and upper slopes. 
Plummer (1899, p. 133) observed that "ancient 
burns, of which traces still remain in the stand- 
ing timber, cover probably 40 to 50 percent of 
the [Mount Rainier Forest] reserve, but being re- 
stocked with trees of large size cannot be called 
burns." We are not yet prepared to relate fire 
frequency or burning characteristics to the 
various forest types. Evidence for multiple fires 
can be found in most drainages at MRNP. In the 
Ohanapecosh Valley, for instance, fires evidently 
took place about 260, 450, and 670 years ago, as 
suggested by three widely occurring age classes 
of Pseudotsuga menziesii . Charred bark on sur- 
vivors of 450 age class can be attributed to the 
fire about 200 years later. Evidence of repeated 
fires can also be found in places where charcoal 
occurs in situ both above and below the tephra 
W, a pyroclastic deposit about 450 years ago 
(Mullineaux 1974) . Many younger stands were 
initiated by white man's fires. The Sheepherder 
Burn of the White River (oldest trees dating about 
118 years) is a good example. The Cowlitz Valley 
appears to be the most extensively burned within 
the last 150 to 200 years. Plummer (1899) re- 
ported "great burns" of the Cowlitz (possibly out- 
side the Park boundaries) in 1841 and 1856 and 
reburns at intervals around 1889-1899. 

Avalanches are another common disturbance, 
affecting forests at mostly high and intermediate 
elevations (Table 1) . Their extent is particularly 
apparent on air photographs. The start zone, 
trimline, and runout zone can be contrasted to 
more irregular patterns produced by fire. The 
aerial extent of successional vegetation of ava- 
lanche paths is not proportionately reflected in 
our sample of 400 plots. Vegetation composition 
of avalanche areas has been studied in the North 
Cascades by Cushman (1976) . She found species of 
Alnus , Acer circinatum , and Pteridium aquil inum 
to be important dominants on south-facing tracks. 

Lahars (Crandell 1971) are another widespread 
agent of disturbance. The 1947 Kautz lahar is a 
conspicuous example of almost primary succession 
on newly deposited surface. Root burial produced 
total mortality of the tree overstory. Dominant 
serai shrubs and herbs include Alnus rubra, A. 
sitchensis (at higher elevation), Salix , Pteridium 
aquilinum , Epilobium angusti folium , and Anapha 1 is 
margaritacea . Another lahar surface adjoining the 
1947 lahar but older than 450 years has forest of 
the Silver fir/Alaska huckleberry type that we 
aged at about 590 years. As we were unable to 
find any charcoal residues that might have indi- 
cated a fire history, the forest might be first 
generation primary succession. The surface might 
correlate to the Electron lahar about 600 years of 
age (Crandell 1971) . This forest differs little 
from stands of Silver fir/Alaska huckleberry in 
the Ohanapecosh Valley of about the same age but 
evidently fire-originated. Lahar deposits as soil 
material differ little from recent alluvium, pyro- 
clastic deposits, or colluvium in affecting old- 
growth vegetation composition. 

Alluvial and glacial processes also cause 
forest succession. A common tree invader of new 
river bars is Populus trichocarpa . Glacial out- 
wash torrents and meltwater deposits destroy old 
banks and cause primary succession on new sur- 
faces. Our plot 356 (Table 3) thus originated. 
Glacial advances create forest disturbance, and 
recessions initiate succession on primary surfaces 
(Sigafoos and Hendricks 1972) . Strong trimlines 
can be seen along the Nisqually River near the 
highway bridge. Plot 375 (Table 3) below that 
trimline occurs on morainal drift exposed around 
1840 (Crandell 1969) . 

Windthrow was extensive enough in several of 
our plots to qualify as a major disturbance. But 
we could not ascertain whether the blowdown itself 
was the immediate agent of succession or followed 
upon another event such as root rot, tree kill by 
insects, or fire. 

Biotic influences on forest succession at MRNP 
are important and varied. Effects can be local 
or widespread. Examples include beetle kill of 
Pinus monticola during the 1960s, beaver impound- 
ments creating red alder communities, pockets of 
root rot (Phellinus weirii) , tree girdling by 
foraging bears, and feeding behavior of large 
herds of elk. Elk activities are apparent in many 
of our plots in the Ohanapecosh, Cowlitz, and 
Mowich River sectors. The migrating herds have 
seasonal impacts in several of the forest types. 
Recurrent usage of the same areas bv elk could 
result in significant shifts of vegetation domi- 
nance . For example the greater abundance of 
devil's club (Oplopanax horridum) in the Carbon 
River (no elk) than at Ohanapecosh Valley (numer- 
ous elk) may reflect both climatic and elk-use 
factors, since this is a favorite browse plant. 


Figure 4 shows the distribution of tree ages 
in samples from Cowlitz, White, and Ohanapecosh 
River drainage systems. Our selection of trees 
for age determination was intended to provide an 
estimate of time since disturbance, so that the 
histograms are a sample of only the older, serai 
trees. Douglas fir was usually present in mcst 
disturbed forests at low and intermediate eleva- 
tions. Other serai species included noble fir 
and western white pine. At higher elevations 
we sometimes had to resort to mountain hemlock, 
Alaska yellow cedar or silver fir, recognizing 
that these could be early- to mid-seral as well 
as late serai or climax. 

These age patterns reflect a 
turbances discussed above, but 
The histogram for the Ohanapeco 
three modes. Peaks occur at 22 
years. As seedling establishme 
years after the disturbance event 
sonable to speculate major dist 
at about 275-300, 750-800, and 
ago. Of course other times of 
are likely too (e.g., specimens 
aged at 450 years or noble fir 
and possibly our sample has not 
vealed these periods. 

variety of dis- 
principally fire. 
sh drainage shows 
5, 675, and 975 
nt can lag many 
,it is not unrea- 
urbance periods 
1000 or more years 
forest disturbance 

of Douglas fir 
at 310 years) , 

adequately re- 

The Cowl 
in the past 
major modes 
and 1000 ye 
Plummer's ( 
another seq 
drainage . 
at 75, 475, 
probably ag 
white man's 
peaks may s 
tioned by P 

itz drainage ha 

two centuries 

at 75 and 175 

ars . The 75-ye 

1899) burns of 

uence is appare 

Ages of serai t 

and 775 years. 

ain reveals fir 

activity, espe 

r Burn" on Crys 

how periods of 

lummer (1899) . 

s been heavily 

Our histogram 
and minor peaks 
ar peak may ref 
1896 and earlie 
nt in the White 
rees there show 

The youngest 
es contemporane 
cially the so-c 
tal Mountain, 
"ancient burns" 

at 675 


r. Yet 


ous with 


The older 

Regard Figure 4 as only a first approximation 
to disturbance periodicity in forests at MRNP. 
The samples are not necessarily proportionate to 
the actual aerial extent of the various distur- 
bance types and periods within forest populations. 
As stands mature and mortality reduces the old tree 
record, the probability of sampling the oldest 
trees of the earliest serai cohort diminishes. 
Multiple disturbances long ago may also be impos- 


sible to resolve from sparse records of the old- 
growth survivors. 


Figures 5 and 6 were made by arranging by age 
plots of the Silver fir/Foamf lower and Western 
hemlock-Silver f ir/Oregongrape forest types. One 
problem making chronosequences is recognition of 
early members when continuously intergrading sam- 
ples up to old-growth climax do not exist. Our 
samples had a gap in the record about 400-500 
years ago. But the plots before and after this 
gap were f loristically similar enough to rea- 
sonably assign to the appropriate chronosequence . 

We detect some rather uncertain understory 
trends in the Silver fir/Foam flower forest type 
(Fig. 5). Foam flower (Tiarella unifoliata) itself 
is rather variable, but has higher cover percentages 
in plots over 500 years than those younger than 
400 years. Bracken fern does not persist after 
about 200 years. Similarly both twinflower (Linnaea 
borealis) and bunchberry (Cornus canadensis ) seem 
more associated with younger plots of the noble fir 
successional stage. In the Western hemlock-Silver 
f ir/Oregongrape forest only bracken fern revealed 
any trend: it occurred (from 1 to 10 percent 
cover) in only four plots under 125 years of age. 

T ' r 


•— L'.'Efa 

WHITE RIVER i!5 T»tts 

FIGURE 4. Sampled age distribution of oldest 
trees from various forest stands in three river 
drainages . 

Rapid understory recovery after disturbance in 
two of the wetter forest types is also suggested 
in Table 3. Plots within each type were compared 
by calculating percentage similarity from 14 domi- 
nant understory species. Comparisons included 
young plots (secondary succession after fire and 
primary succession on new geological substrates) 
with old-growth. We regard vegetation between 
plots to be similar if calculated similarities ex- 
ceed about 25 percent. Similarities of each plot 
with others of the same forest type often exceeded 
25 percent regardless of forest age or disturbance 
type. We discerned no striking or consistent 
difference in dominant understory vegetation 
whether 100 or 500 years after the disturbance. 

' 1 1 




- V 

Tiorgllo unifolioto 
O O 
Pleridlum aquilinum 
Linnaea bona lis 

Cornus canadensis 


~ ■ & *l 

J I 

200 400 


FIGURE 5. Cover of selected herbs in relation to 
plot age, Silver fir/Foamf lower forest type. 

Tree composition (Fig. 6) is the most indica- 
tive state of the sere . Noble fir usually dominates 
the Silver fir/Foam flower habitat during the first 
several centuries after fire. (All plots of fig- 
ure 6 were apparently fire initiated . ) Alaska yel- 
low cedar, western hemlock, Douglas fir, and sil- 
ver fir may also be serai, however. But after 500 
years silver fir is usually the leading and some- 
times the exclusive dominant. By contrast western 
hemlock or Douglas fir are several dominants after 
fire in the Western hemlock-Silver fir/Oregon- 
grape type. The proportion of Douglas fir grad- 
ually declines, with replacement by the hemlock. 
Silver fir achieves very small share of the canopy 
dominance only in the oldest plot. Tree regen- 
eration patterns differ in the two forest types. 
Western hemlock has continuous regeneration 
pressure throughout the time span of the Western 
hemlock-Silver f ir/Oregongrape chronosequence, 
but decays to minor status in old growth of the 
Silver fir/Foamf lower type. In both chrono- 
sequences silver fir is the eventual (after about 
200 years) regeneration dominant. 











i j 

1 ' ( 1 
i i 

1 4BAM I ' 

! i - 


Y " x 


tshe! 1 ! . ! 





i! ,! 




! 1 ! 

! ■ ■ 

; il i !i " 




1 ' ■ 



' flBOM 

■ 1 . 




J I 



I i - 


i A\\ 

1/ llll 

1 1 

1 ' 1 
1 ll 

i in ! 

M 1 

~~- TSHE 

i 1 1 

ii i ■ 


~~t i i * 

200 400 



FIGURE 6. Tree chronosequences in two forest types. 
Canopy dominants are Douglas fir (solid bars) , wester 
Hemlock (short dashes), noble fir (long dashes), and 
silver fir (dash-dot bars) . Regeneration is silver 
fir (ABAM) , western hemlock (TSHE) , and Douglas fir 
(unlabeled curve) . 


TABLE 3. Comparison of dominant understory composition (computed as percentage 
similarities) between young and old-growth forests in two forest types. 

Forest Type 







ears BP) 






New substrate 


18, 21 






44, 45 



i |A 


44, 42 




2] , 

45, 42 







30, 39 




New substrate 


57, 27 





57, 30 





27, 30 

1/ QA = Quaternary alluvium, QO = Osceola mudflow, QG = Garda drift 
TBR = Tertiary volcanic rock. 

2/ Each plot compared to others of the same forest type sequenced 
from top to bottom of column 2. 

We feel that the first few decades after dis- 
turbance is the period of greatest vegetation 
contrast between young and old stands. After 
about 100 years in mesic forest types of valleys, 
draws, or lower slopes the sere has assumed most 
of the attributes of climax. Successional trends 
seem slower, however, on drier or more exposed 
sites such as Backbone Ridge or upper elevations 
at Sunset Park where forest closure and stabilized 
understory composition may require centuries. 


We have summarized much of our present state- 
of-knowledge concerning forest ecosystems at MRNP. 
Our efforts are now directed to better resolving 
the mature forest types with computer help and 
mapping these and their serai derivatives from 
air photo information. More analysis of succession 
is needed in some of our hot, dry as well as high 
elevation forest types. We need better resolution 
of the frequency, extent, and nature of natural 
and man-caused forest disturbances at MRNP. These 
studies should help provide park managers some in- 
sights into ecological processes that have brought 
about in past centuries the complex forest mosaics 
existing at Mount Rainier today. Land use prac- 
tices both within MRNP and outside the Park may 
or may not disrupt or modify certain of those pro- 
cesses. The ability to predict forest responses 
on both short and long term should give manage- 
ment necessary perspectives for the job of both 
utilizing and preserving the forest ecosystems. 


CRANDELL, D. R. 1970. Postglacial lahars from 
Mount Rainier Volcano, Washington. Geol. Surv. 
Prof. Paper 677, iv, 75 p., map. 

. 1969. Surficial geology of Mount 

Rainier National Park, Washington. Geol. Surv. 
Bui. 1288, vi, 41 p., map. 

CUSHMAN, MARTHA J. 1976. Vegetation composition 
as a predictor of major avalanche cycles, North 
Cascades, Washington. M.S. thesis, Univ. 
Washington, Seattle, vii, 99 p. 

1974. A preliminary classification of forest 
communities in the central portion of the 
western Cascades in Oregon. US/IBP Coniferous 
For. Biome, Bui. 4, viii, 123 p., Univ. Wash- 
ington-AR-10, Seattle. 

FRANKLIN, J. F. and C. T DYRNESS. 1973. Natural 
vegatation of Oregon and Washington. USDA 
Forest Serv. Gen. Tech. Rep. PNW-8, viii, 417 p 

. W. H. MOIR, G. W. DOUGLASS, and C. 

WIBERG. 1971. Invasion of subalpine meadows 
by trees in the Cascade Range, Washington and 
Oregon. Arctic & Alpine Res. 3:215-224. 

, C. T. DYRNESS, and W. H. MOIR. 1970. 

A reconnaissance method for forest site classi- 
fication. Shinrin Richi XII:1-12. 

and R. G. MITCHELL. 1967. Succes- 

sional status of subalpine fir in the Cascade 
Range. USDA Forest Serv. Res. Paper PNW-46, 
16 p. 

HOBSON, F. D. 1976. Classification system for the 
soils of Mount Ranier National Park. M.S. 
thesis, Wash. State Univ., Pullman, 79 p. 

MULLINEAUX, D. R. 1974. Pumice and other pyro- 
clastic deposits in Mount Rainier National Park, 
Washington. Geol. Surv. Bui. 1326, viii, 83 p. 

PLUMMER, F. G. 1899. Mount Rainier Forest Re- 
serve, Washington. Geol. Surv. 21st Rep. 
1900[1901], Pt. 5:81-143. 

SH1MWELL, D. W. 1971. The description and clas- 
sification of vegetation. Univ. Wash. Press, 
Seattle, xiv, 322 p. 

SIGAFOOS, R. S. and E. L. HENDRICKS. 1972. 

Recent activity of glaciers of Mount Rainier, 
Washington. Geol. Surv. Prof. Paper 387-B, 
vi , 24 p. , 6 maps . 


R. W. Fonda 1 


Timberline is a fascinating ecotonal area, 
where there is a pronounced shift in physiognomy 
from trees to meadow f orbs , grasses, and dwarf 
shrubs. This area is created by a complex inter- 
play of factors that shift as elevation and expo- 
sure increase on a mountain. 

Many researchers have tried to explain the 
causes of timberline (Table 1) . Many scientists 
and lay persons have favored low temperature as the 
limiting factor. Others have settled on wind, both 
in winter and during the growing season, as a 
limiting factor, and snow has often been cited. 
One general theme that is evident among the inves- 
tigators listed in Table 1 is that Americans have 
described the forest communities at timberline 
and have generally speculated on the climatic con- 
trols, while Europeans have selected individual 
dominant species and have looked intensively at 
physiological responses to certain factors, in 
the laboratory. In every case, the studies have 
been characterized by a common theme: all have 
searched for the single limiting factor causing 
timberline . 

I hypothesize that the vegetative pattern at 
timberline is the result of the interaction of 
topography, exposure to wind, snow accumulation, 
timing of snowmelt, soil moisture, air temperature, 
and soil temperature. A principal result of this 
interplay is that a site with neither too much 
nor too little snow, nor too little water, sup- 
ports trees at the limits of their distribution 
in the Olympic Mountains. 


Since 1971 my students and I have studied the 
hypothesis shown in Fig. 1, and we are continuing 
to test all the elements of this hypothesis. Our 
work in the Olympics has centered on Eagle Point 
(1829 m) , a prominence along the Hurricane Ridge 
complex in the northeastern portion of the park. 
Here, timberline forests of subalpine fir (Abies 
lasiocarpa) contact alpine and subalpine meadows. 
Furthermore, these trees encounter the entire 
range of environmental conditions described in 
Fig. 1. Sites with low wind, heavy snowpack, and 
late snowmelt are available, as are sites with 
high wind, light snowpack, and early snowmelt. 
The snowmelt pattern was established by Canaday and 
Fonda (1974). In brief, there are four meadow 
communities plus the treeline. True alpine vege- 
tation is snowfree during winter; a mesic grass 
community occupies the outer portions of the mead- 
ow, and it is released from snow in early June; 
next is a tall sedge community that is released 
from snow in July; the center of the snowfield 
supports a dwarf sedge community, released from 
snow in August. The trees at Eagle Point are free 
of snow by mid-July, about the same time as the 
tall sedge community. The band of trees that our 
intensive work is conducted upon is bordered by a 
mesic grass community to the windward side, and a 
tall sedge community to the leeward side. 

Biology Department, Western Washington State 
College, Bellingham, Washington 98225. 

Eagle Point appears to be almost a constant en- 
vironment. Although the site is basically a con- 
vexity, it behaves as a concavity because of the 
line of trees forming timberline. There is a 
limit to the maximum snow accumulation, because 
any snow that accumulates above the influence of 
the trees will be blown off the ridge. Conse- 
quently, year after year about 4-5 m of snow 
piles into the center of this meadow. This tapers 
to about 1 m over the tall sedge community, and 
to less than 1 m over the mesic grass community. 
Rate of snowmelt is constant, if the summer is not 
overly cloudy and rainy. 

The summer of 1972 characterizes the typical 
climatic relationships during the growing season 
at Eagle Point. Summers are generaly sunny, warm, 
and dry. Daily mean VPD is seldom greater than 
5 mm Hg , with a mid-day maximum of 9-11 mm Hg. 
Lowest hourly values occur between 0200-0700 hrs . 
Most days have over 500 langleys of radiation, 
and mean daily air temperature consistently ex- 
ceeds 10 C, except during storms. Mean daily soil 
temperatures do not reach 10 C. The treeline sta- 
tion at Eagle Point has a Temperature Growth Index 
(TGI) of 27. Waring etal. (1972) found that moun- 
tain hemlock stands near timberline in the 
Siskiyou Mountains have a TGI=30, so that our data 
appear to reflect the typical temperature regime 
for such stands in the Pacific Northwest. (See 
Cleary and Waring 1969, Waring 1969, and Waring 
etal. 1972, for a definition and use of TGI). 

Figure 2 shows soil moisture values for four 
stations at Eagle Point in 1972. Note that soil 
moisture regime at the treeline station is most 
like the mesic grass/tall sedge ecotone, and that 
beginning in August the mesic grass community is 
significantly drier than the trees. The dwarf 
sedge community is significantly wetter than the 
treeline station. This is certainly enough mois- 
ture to support trees, but the August snow release 
mitigates against the tree growth form. 

Most, but not all, summers have these kinds of 
environmental relations. The pathways depicted 
in Fig. 1 need not obtain every year for trees to 
be limited in their distribution. The summer of 
1976, for example, was atypical, because it was 
wetter and cooler, with many cloudy days. Most 
days in 1976 had less than 500 langleys, and mean 
daily air temperature frequently stayed below 
10 C. Almost the same total precipitation was 
recorded during the summers of 1972 and 1976: 
94 mm and 90 mm. But, in 1972 precipitation was 
almost equally distributed among the summer 
months, whereas in 1976 about 60% fell during 
August. The principal difference between the 
summers of 1972 and 1976 was the almost constant 
cloud cover of 1976, which delayed snowmelt and 
kept sites from drought. Subalpine fir was sub- 
jected to water stress in 1972, but not in 1976. 


Basins and leeward concavities, which accumu- 
late much snow during winter, mark the limits of 
timberline along one end of the environmental 
gradient at Eagle Point. The problems of too much 







FIGURE 1. The interaction of environmental factors and plant physiological responses that 
produce timberline and meadow vegetation in the Olympic Mountains, Washington. 





tt T 

■-> O 









/•I - 







20 30 10 20 30 10 20 30 10 20 

FIGURE 2. Soil moisture trends at Eagel Point during summer 1972. Each soil moisture 
record was begun when the soil thermistor block leads were exposed by snowmelt. 


TABLE 1. Important environmental controls at timberline and associated effects. 

Low temperature 

Winter wind 


Duration of non-photosynthetic period 
Photosynthesis end production 
Freeze-thaw cycles in cells 

Remove snow cover 
Cause snow accumulation 

Frost drought 
Frost hardiness 

Invc s t i gat or = 

Tranquil lini 

Tranquillini ; Kooney; Daubenmire 

Tranquil lini 


Kronfuss; Tranquillini; Cai 

U Fonria 
Tranqui 1 lini 

Summer wind Increase transpiration 

Lower leaf water potential 

Snow Protection from freeze-thaw 

Compaction of trees 

Block winter transpiration 

Summer drought Excessive transpiration 
Internal water stress 
Decrease photosynthesis-respiration 

Fire Succession 

Caldwell; Griggs; Daubenmire; 

Wardle; Klikoff 

Billings; Arno; Frar.klin et el; 

Fonda & Bliss 
Tranqui 11 ini 

Tranqui 1 1 ini 
Kuramoto & Bliss 

Morris; Billings; Griggs; Fonda 
& Bliss 

Micro topography 

Differential snow accumulation 

Several of the above 

TABLE 2. Mean plant water potential (bars) at Eagle Point, 1972; morning measurements, 

7-14 July snowmeit 

16 June snowmeit 

20 May snowmeit 










8 Aug 






14 Aug 





25 Aug 






5 Sep 






snow have been amply doc 
1971, Fonda and Bliss 19 
1974) . Heavy snow damag 
they are able to survive 
and stunted into krummho 
tantly, however, the Aug 
that accumulates several 
in a growing season that 
tree to produce the nece 
tissues for the dormant 
plete a reproductive eye 
perennial tissues above 
which few timberline pla 
stead, meadow species ar 
well adapted to the sequ 
at Eagle Point (Canaday 

umented (Franklin et al 
69, Canaday and Fonda 
es the trees, so that if 
at all they are twisted 
lz forms. More impor- 
ust release of a site 
meters of snow results 
is too short for the 
ssary wood, harden off 
season, and perhaps com- 
le (Fig. 1) . Wood and 
ground are a luxury for 
nts have a strategy. In- 
e favored, and they are 
ential release of snow 
and Fonda 1974) . 

Snow does not accumulate on windward sides of 
ridges in winter, and frost drought can be a 
problem in spring (Fig. 1) . On windward sites 
high winds, high daytime temperatures, and low 
soil moisture combine to act against trees during 
the growing season. Because precipitation is 
so light at Eagle Point, the principal recharg- 
ing agent for soil moisture is snowmelt. Light 
snow results in a poor soil moisture recharge, 
and a possibility of drought during the growing 
season (Fig. 1). The growing season at timber- 
line is short. Based on phenological observations 
and dendrometer measurements, subalpine fir at 
timberline begins to show bud swell and an in- 
crease in trunk diameter around 1 July. These 
trees cease growth between 12-22 August, although 
cones continue to mature after that date. About 
300 m lower on the slope, subalpine firs in the 
closed forest begin to break dormancy shortly 
before those at timberline, but they continue to 
grow for about four weeks after those at timber- 
line have stopped. 

One factor that contributes to cessation of 
growth at timberline is low water potential in 
August (Table 2) . Data on water potential were 
gathered with a pressure bomb, from dawn to early 
morning during August and early September 1972. 
On sites that either support subalpine fir or that 
behave similarly (ie, tall sedge), the water po- 
tential values were slightly less than -10 bars 
(Table 2) . On the windward side of the treeline 
at Eagle Point is a mesic grass community. 
There are no subalpine firs growing there, so that 
dwarf juniper, a common component of the commu- 
nity, was used for water potential measurements 
(Table 2) . The values for juniper in the mesic 
grass community are lower than values obtained 
for subalpine fir at the same time, which indi- 
cates that low soil moisture may be a barrier to 
subalpine fir invasion. 

A water potential of -10 bars is a 
division point, because subalpine fir 
tolerate internal water potentials as 
bars. Once the values drop below -10 
photosynthesis, respiration, and tran 
decrease sharply for subalpine fir (P 
1973) . Early morning values less tha 
indicate that the timberline trees ar 
critical point throughout the day dur 
and that growth should cease. Consis 
our observations, Waring etal. (1972) 
forest stands near timberline did not 
potentials less than -10 bars. 


Figure 1 summarizes the main environmental 
problems facing subalpine fir at timberline in 
the Olympic Mountain. These trees must adapt to 
many possible limiting factors, all of which 

n important 
can easily 
low as -10 

, however , 



n -10 bars 

e below that 

ing August, 

tent with 
found that 
show water 

have extreme expression in a space of a few 
meters. At Eagle Point these factors combine to 
limit the effective growing season for subalpine 
fir. Consequently, timberline represents a point 
where trees show an early cessation of growth and! 
low biomass accumulation, in addition to other 
physiological stresses. My students and I have 
gathered data on some of these relationships. Ot: 
present work focuses on summer water balance, es- 
pecially stomate behavior, phenology, and growth 
response. Once we fully understand the water 
status of subalpine fir during summer drought, we 
hope to move to the next set of related responses* 
photosynthesis/respiration rates and productivity.'! 
at timberline. 


CANADAY, B. B. and R. W. FONDA. 1974. The in- 
fluence of subalpine snowbanks on vegetation 
pattern, production, and phenology. Bull. 
Torrey Bot. Club 101:340-350. 

CLEARY, B. D. and R. H. WARING. 1969. Tempera- 
ture: collection of data and its analysis for 
the interpretation of plant growth and distri- 
bution. Can. J. Bot. 47:167-173. 

FONDA, R. W. and L. C. BLISS. 1969. Forest vege 
tation of the montane and subalpine zones, 
Olympic Mountains, Washington. Ecol . Monogr . 

C. A. WIBURG. 1971. Invation of subalpine 
meadows by trees in the Cascade Range, Washing 
ton and Oregon. Arc. Alp. Res. 3:215-224. 

PURITICH, G. S. 1973. Effects of water stress 
photosynthesis, respiration, and transpiratior 
of four Abies species. Can. J. For. Res. 3: 

WARING, R. H. 1969. Forest plants of the eastei 
Siskiyous: their environmental and vegetation 
distribution. Northw. Sci. 43:1-17. 

1972. An environmental grid for classifying 
coniferous forest ecosystems, p. 79-91. in: 
J. F. Franklin, L. J. Dempster, and R. H. War- 
ing (eds.). Research on coniferous forest ecc 
systems - a symposium. USDA For. Serv., Port- 
land, OR. 




Donald B. Lawrence' 


An alternative title could have been: "Some 
results of sixty years of research by the Univer- 
sity of Minnesota Department of Botany at Glacier 
Bay National Monument." The work was begun in 
1916 by Professor William S. Cooper, who is now 
92 years old. This paper is dedicated to him on 
the 60th anniversary of the first expedition. 
That was long before there was a National Monu- 
ment. It was through Cooper's efforts that the 
Monument was established in 1924 (Cooper 1956). 
I shall emphasize the wonderful opportunities 
there for studying patterns and processes of phys- 
iographic ecology. Primary succession has 
occurred following catastrophic glacial recession 
in the main bay (Figs 1-3) from maximum ice posi- 
tions of the Little Ice Age, or Neoglacial as it 
is otherwise known, about 225 years ago, or 
roughly 1750 A.D. Secondary succession can be 
studied at Lituya Bay on the west coast of the 
Monument (Figs. 4-5) where giant waves generated 
by huge rockslides, most recently in 1958, swept 
away the forest but not all the soil from the low- 
er slopes of that bay. Secondary succession at 
Lituya Bay has. attained about the same develop- 
ment in 15 years as primary succession has in 
50-100 years on the shores of Glacier Bay, both 
starting mainly from seeds and spores. This is a 
maritime region of relatively mild moist climate 
and there is no permanently frozen ground. 

Generous assistance of National 
staff has been received in recent 
cially by former Park Superintende 
and by Naturalists Gregory P. Stre 
Bruce Paige, and by Captain James 
Motor Vessel Nunatak, and mate Eug 
My wife, Elizabeth, has cooperated 
Several other colleagues have help 
Lloyd Hulbert, Robert Crocker, Rod 
Roland Schoenike, William Reiners, 
Gordon Nichols, and James Taylor; 
were members of the staff of the N 
Service of New Zealand at the time 
Field invited me to participate in 
my first to Glacier Bay; he has be 
help and inspiration, as has Coope 

Park Service 
years, espe- 
nt Robert Howe , 
veler, and 
Sanders of the 
ene Chaffin. 

in the work. 
ed, especially 
erick Sprague, 

Ian Worley, 
the last two 
ational Park 
William 0. 

his 1941 trip, 
en of continuing 
r over the years. 

Cooper selected Glacier Bay 
he wanted to gain a better unde 
vegetation became established i 
ing recession of the continenta 
Pleistocene. His dream of this 
for understanding the developme 
landscape has been so successfu 
sible to predict that fossil ev 
avens (Dryas) would eventually 
sota sediments (Schoenike 1958) 
was published several years bef 
leaves were actually discovered 
Minnesota (Watts 1967) where th 
no longer lives today. 

for study because 
rstanding of how 
n Minnesota follow- 
1 ice sheets of the 

study as a model 
nt of the Minnesota 
1 that it was pos- 
idence of alpine 
be found in Minne- 
This suggestion 
ore fossil Dryas 

in 1961 . n northern 
is genus of plants 

Department of Botany, University of Minnesota, 
St. Paul, Minnesota 55108. 

I air. glad that the early observations of this 
study were made before there was much visitation 
by tourists because the validity of the results 
has depended on absence of human influences. 
People are recognized dispersers of plants into 
natural areas (Ridley 1930) . Their activities in 
altering the biota and pattern and rate of succes- 
sion have been dramatic in some national parks, 
as for example in Hawaii, and in New Zealand. In 
conducting this research I have always been care- 
ful to begin the observations at the youngest 
sites, wearing carefully cleansed clothing, and to 
proceed to older and older sites so that I would 
not myself be altering the progress of the succes- 
sion (Cooke and Lawrence 1959: 531-532). Much 
more care is needed in this respect by visitors 
who go ashore especially in the upper bay today. 


We may define primary succession as the process 
of development of ecosystems on raw parent material 
that has no soil profile and contains no viable 
remains of organisms; community establishment must 
commence anew by immigration from other areas. 
This region is perhaps the most interesting por- 
tion of the earth's surface for the study of 
history of glaciers and the development of vegeta- 
tion following recession of ice. This is true be- 
cause of the extraordinary rate of recession since 
the position of maximum advance about the middle 
of the 18th century. Since then there has been 
recession of 105 km (65 mi) to the faces of Grand 
Pacific Glacier and Johns Hopkins Glacier at the 
heads of the west arm of the bay, and of 8 km 
(48 mi) to the Muir ice front at the head of Muir 
Inlet (see Fig. 1 based on USDI NPS map) . This 
process has exposed new fiord systems, U-shaped 
valleys, and numerous habitats including bare rock, 
till, outwash, ponds, and many erect stumps rang- 
ing in radiocarbon age from 7000 yrs near the 
head of Muir Inlet to about 300 yrs near the mouth 
of the bay (Lawrence 1958) . 

The map of Glacier Bay (Fig. 1) shows known 
positions of the ice fronts (solid lines) , and 
estimated positions (broken lines) based on a 
variety of sources. We are indebted to John Muir 
for a sketch (Fig. 3 top) of the position of "The 
Pacific Glacier" at the head of the main bay in 
1879, when it extended northeast from the south- 
ern part of Russell Island to the mainland (Muir 
1895; 243, and Bohn 1967: 48). Early positions 
of the front of Muir Glacier are based on a de- 
scription by John Muir in 1880, on photographs by 
G. F. Wright in 1886, and surveys by H. F. Reid 
in 1890 and 1892 (Gilbert 1904: 21, Fig. 10). 
More recent positions are known mainly from sur- 
veys by Field (1947), and Goldthwait et al. (1966). 
The known history of these glaciers and other moun- 
tain glaciers of the whole northern hemisphere has 
been assembled in detail by Field (1975). Time 
of recession from the mouth of the bay about 1750 
is estimated from growth layer counts from basal 
cores of oldest Sitka spruce trees by Cooper on 
the terminal moraine near the east shore, and our 
own there and on the west shore where seedling es- 
tablishment occurred a few years earlier. 







\ ■• I, > 

iv w.icoo*f« y^ /& L 

% icc r«o«T / y 

HI «IVCN fill ' 


FIGURE 1. Map of Glacier Bay National Monument showing history of glacial recession, 105 km (65 mi) in 
the main bay, and 80 km (48 mi) in Muir Inlet in the last 225 years since the Neoglacial (Little Ice 
Age) maximum when the glacier reached Icy Strait. Note location of Cooper Plot #1 near the head of the 
main bay where ice front stood in 1879. Lituya Bay is at the left between the two main areas of pre- 
Wisconsin refugia. 


The general progress of vegetation develop- 
ment on till can be demonstrated most readily on 
the shores of Muir Inlet (Fig. 2) because there 
have been large portions of glaciers, cut off 
from sources of supply, gradually melting away 
on relatively level surfaces (Fig. 2A) . Their 
margins have been observed almost annually so that 
one can know quite precisely where the ice edge 
was in a particular year and be sure that no vas- 
cular plants were growing on debris overlying 
buried ice. Thus the time relations of develop- 
ment of vegetation and soil (Crocker and Major 
1955, Ugolini 1966) can be documented beginning 
with soil parent material of age zero years. 

Whence came the plants to occupy the new 
surfaces ? -- We have discovered by observation 
that reproductive parts of plants, at least of 
green plants, cannot survive beneath glaciers 
through even a "Little Ice Age" of only a few 
hundred years. The process of vegetational de- 
velopment must begin with immigration of mobile 
seeds and spores from adjacent areas of living 
vegetation blown in by wind or carried on the 
pelts and in the digestive tracts of wild ani- 
mals, or on the clothing of human visitors. 
There are two main sources of spores and seeds 
at Glacier Bay. One is the f loristically rich 
undisturbed lowland forest and muskeg beyond the 
Neoglacial (Little Ice Age) ice limit near the 
mouth of the bay; the other is above the Neo- 
glacial high ice limit, mainly alpine tundra with 
scattered mountain hemlocks on the mountain slopes 
that flank the bay. The locations of these may 
be seen on the map of Figure 1; they are based 
on the map of Streveler and Paige (1971, Fig. V). 
It will be noted that the vegetational trimline 
which was carved by the glacier as it advanced 
down the bay, lies at sea level in the first 26 km 
(16 mi) northward from the mouth of the bay to 
Beartrack Cove, rises northward rather steeply 
along forested slopes in the next 13 km (8 mi) 
to 610 m (2000 ft) east of Sandy Cove, then rises 
much more gently mainly above tree line for the 
next 35 km (22 mi) to 760 m (2500 ft) on the 
slopes of Red Mountain east of Muir Inlet. 
These elevations were measured by Rossman and 
Loken (1955 personal communication) from a small 
plane with altimeter and pocket level. 

Northwestward from the isolated tundra area 
on the promontory that separates Glacier Bay from 
Muir Inlet, and northwestward from the ridge just 
south of the mouth of Geikie Inlet no relict 
areas of high vegetation have been found above the 
Neoglacial high ice line from which seeds and 
spores could be carried to the deglaciated areas 
about the head of the main bay. However, the 
mountains above the high ice line are almost un- 
explored, and isolated refuge areas may yet be 
discovered. Cooper (1942) described an isolated 
colony including 15 species of seed plants at 
elevations of 1250 m (4100 ft) to 1830 m (6000 
ft) on an eastward ridge of glacier-clad Mt 
Bertha, a peak of 3100 m (10,800 ft) which juts 
up from the northern part of the Brady Icefield 
(Fig. 1) . He considered this to be a pre- 
Wisconsin relict community. From these facts it 
becomes obvious that the distance that had to be 
traveled by seeds and spores was very short 
when the ice was receding from the mouth of the 
bay about 1750 A.D., but only a few of the spe- 
cies present in the mature forest would have 
pioneering capabilities. At the head of the main 
bay, and especially in the very isolated area 
along the shores of Johns Hopkins Inlet, the 
sources of plant reproductive parts are far a- 
way. But as the glacier has receded gradually 
up the bay and up Muir Inlet, plants with 

greatest dispersal mobility and capability for 
developing promptly to the reproductive stage on 
raw substrata following ice recession have be- 
come established in vacated areas, leaving a bare 
zone of varying width near the melting ice. Be- 
cause the ice sheet has receded vertically as well 
as horizontally, the upper slopes have naturally 
become vacated before the adjacent lower ones so 
that with passage of time there was less distance 
to be traveled by the disseminules of pioneering 
plants that had reached reproductive maturity in 
order to colonize the new bare areas near sea 
level. Thus it may be seen that the progress of 
succession moved across the landscape in a wave- 
like motion from the older to the younger sur- 
faces gradually filling in the barren U-shaped 
belt left along the tip and edge of the receding 
ice . 

what kinds of plants pi oneer .'--Several kinds of 
habitats are left as the ice recedes, in addition 
to marine surfaces which we cannot consider here. 
The soil parent material in a number of places 
is a mixture of gravel, sand, and silt, but gen- 
erally very little clay, a mixture of size classes 
we call loam. After the glacier has melted there 
is no permafrost. Before any plant cover develops, 
freezing and thawing near the surface in the first 
few years bring the uppermost gravel to the top 
and result in a pebble pavement. It is in the 
cracks between pebbles and on the pebble surfaces 
that plants begin to establish. In other places 
where gravel is absent a silty or sandy surface 
occurs, on which a "black crust", actually very 
dark green, develops within a very few years. 
Worley (1973) has made a careful study of this 
and has been able to distinguish three types: 1) 
Lophozia Mat, a minute leafy liverwort, in some 
places mixed with various algae and mosses; 2) 
Cyanophyta Mat, consisting of a tough layer of 
bluegreen algae, the cells Of which are imbedded 
in a gelatinous sheath forming a rubbery layer; 
in some places Lophozia and mosses are also pres- 
ent; 3) Lichen Mat, consisting of colonies of 
small bristle-like lichen Lempholemma radiatum 
2-4 mm tall, and another unidentified lichen. 
These crusts are constructive in that they have 
a stabilizing role; they greatly restrict the 
movement of soil particles under the effect of 
rain drops, solif luction, freezing and thawing. 
In places, movement of the whole mat occurs on 
the steeper slopes "not unlike a carpeting slid- 
ing upon a flooring", according to Worley, making 
the directions of motion obvious by the patterns 
of convex flowlines that develop. Worley con- 
sidered that these mats are not an essential pre- 
cursor for any later successional community, but 
we shall reconsider this because, as he points 
out, they may absorb heat due to their dark color, 
as well as consolidate the upper 1-2 cm of the 
substratum. Furthermore, he found that the varie- 
gated scouring rush (Equisetum variegatum) "is 
common in or adjacent to the crust habitats, its 
rhizomes often underlying the mat." We shall 
see that this member of the fern allies forms an 
important and long-surviving component of the 
developing community, and it probably becomes es- 
tablished in the black crust. Bierhorst (1971a) 
has pointed out that although the minute Equisetum 
gametophytes derived from spore germination often 
die young, they may continue to grow for some 
years with individuals forming discs 3 cm in diam- 
eter, but that they are rarely seen in nature, and 
if seen are often not recognized because of their 
similarity to filamentous algal growths or to 
masses of moss protonemata. It would seem likely 
therefore that these early stages of Equisetum 
occur in the black crust, and further search for 
them would be in order. 


on of Muir Glacier emphasizing 
FIGURE 2. Stages of progress of P-^/^-f^^s' (^s) ! anTr-G Sitka alder (««.) showing 11 
importance of two N-fixing shrubs: B-F alpine avens (Dry , ^ ^ ^ ^^ later> 

importance of two N-nxing shrubs: B"* JJP™^"* Session and 50 years later, 
yrs of change. H-I Muir Point a few years 


Other plants are successful pioneers on the 
pebble pavements mentioned earlier. The moss 
Rhacomitrium canescens and the lichen Stereocaulon 
sp. start on many pebbles. Seedlings of willows 
of several species and black cottonwoods may be- 
come established within the first few years as 
well as Dryas , a mat-forming shrub, and various 
herbaceous plants such as willow herbs of the 
genus Epilobium (Cooper 1923, Decker 1966). It 
is only at this very early stage before other 
plants have formed a dense cover that these spe- 
cies can become established, partly because they 
cannot endure shade and partly because they need 
a mineral surface or some special physical, chem- 
ical or microbial characteristics that are pres- 
ent only at this early stage. This fact has not 
been adequately emphasized heretofore. These 
species have the most mobile disseminules , be- 
ing small, or, in some, being equipped with deli- 
cate plumes or parachutes of hairs which carry 
them with the gentlest breeze. 

Somewhat later, other species with larger seeds, 
such as Sitka alder and Sitka spruce arrive, 
both bearing wings which increase their mobility 
with wind, especially on snow crusted over with 
ice. These species are sometimes accompanied 
within the first decade by an occasional western 
hemlock, also with winged seeds, by two herba- 
ceous members of the pea family, Astragalus 
alpinus and Hedys arum al pinum , and the soapberry 
shrub, Shepherdia canadensis ; these last three 
have large wingless seeds transported mainly by 
birds, and the soapberry often by bears, and per- 
haps all by voles which are very numerous in some 
years. Plant communities that develop close 
above the strand line are salt-tolerant and have 
a very different composition. I shall not con- 
sider them here. 

From what sources come the combined forms of 
nitrogen so necessary to the health and rapid 
growth of ordinary green plants ? — All of the pio- 
neer plants above the strand line except Dryas, 
alder, soapberry, the members of the pea family, 
and those of the black crust, are of a sickly 
yellowish color with very slow growth rate and 
prostrate stature, even the willows, black cotton- 
wood, Sitka spruce and hemlock which are capable 
of rapid erect growth later in life. These are 
symptoms of deficiency of soil nitrogen as has 
been demonstrated (Lawrence 1958) by the spectac- 
ular improvements induced through experimental 
applications of ammonium nitrate to the soil 
around test cottonwood saplings near the head of 
Muir Inlet. Small amounts of available nitrogen, 
enough for minimal growth of pioneer plants with 
lowest nitrogen requirements, may be present in 
the rain water. Another small supply may come 
from the ammonium chloride in many igneous rocks 
(Ingols and Navarre 1952) such as some of those 
here in the raw till. But the major source here 
must come via the process of biological fixation 
of atmospheric nitrogen. The soil is too cold 
for effective fixation of nitrogen by free- 
living bacteria, even by the photosynthetic ones 
known to have that capability. But some blue- 
green algae long known to be able to fix nitrogen 
(Singh 1961) thrive here in quantity within the 
black crust, both independently in the "lyanophyta 
Mat and in symbiosis within the bodies of 
Lempholemma in the Lichen Mat (and also in Stereo- 
caulon lichen). Furthermore, Griggs (1933) in 
his study of the colonization of the Katmai 
volcanic ash farther to the north found that 
liverworts grew on the ash in pure stands, and 
in experimental studies (Griggs and Ready 1934) 
demonstrated that these same liverworts would 
thrive on a medium too low in combined nitrogen 

to support growth of ordinary plants. Thus, it 
seems possible that the Lophozia Mat liverwort 
may possibly fix nitrogen either independently or 
by some obscure symbiotic relationship. 

All vascular plants mentioned above as having 
healthy green color and early rapid growth rate 
have root nodules inhabited by star molds (actinomy- 
cetes) or bacteria, both of which are capable of 
fixing atmospheric nitrogen (Stewart 1966, Becking 
1970) . 

Two of the woody plants mentioned above, the 
Dryas or alpine avens , a member of the rose family, 
and the Sitka alder (Alnus crispa subsp. sinuata), 
belonging to the birch family, have extraordinarily 
rapid growth rates; both are endowed with coralloid 
root nodules in which atmospheric nitrogen is 
fixed symbiotically (Lawrence et al. 1967, 
Lawrence 1958). The Dryas, capable only of hori- 
zontal growth, forms rapidly enlarging disc- 
shaped clones (Fig. 2B, E) that soon coalesce to 
form a continuous blanket (Fig. 2C) over the land- 
scape in many places. While that is going on the 
alder becomes established in open areas where 
there is yet no Dryas mat (Fig. 2F) , and in the 
following decade forms a nearly continuous thicket 
(Fig. 2G) ; subsequently the Dryas dies from shad- 
ing. The alder can become established over a 
longer period even beneath a continuous willow 
canopy, but it seems unable to establish new 
seedlings beneath its own canopy or on the forest 
floor beneath a cottonwood or conifer canopy; 
again, this fact has not been adequately empha- 
sized in the literature. Both Dryas and alder 
have enormous influence on the appearance of the 
landscape as may be seen in Figure 2, and on 
the lives of their associated organisms (Fig. 3) , 
as has been described by Lawrence (1958). Each 
one either kills off the plants beneath it, or 
else stimulates them to very rapid growth. The 
stimulation can result from digestion of the pro- 
tein-rich leaves of the nitrogen fixers by fungi 
to liberate nitrogen in available forms. In addi- 
tion, Bierhorst (1971b: 646) has observed that 
micorrhizal fungi from one plant incapable of fix- 
ing nitrogen can penetrate root nodules of an 
adjacent symbiotic nitrogen fixer to obtain ni- 
trogen compounds more directly. 

By the time 50 years have passed 
of the ice, a dense alder thicket u 
ft) tall had developed at Muir Poin 
from which have emerged the same co 
trees that became established soon 
cession, but which could not grow e 
places for over 40 years, until the 
vided them with adequate amounts of 
rapid erect growth. Thus the "law 
in which each stage is supposed "to 
way for the next stage, and thereby 
its own elimination" , is turned top 
in this case the pioneer cottonwood 
the succeeding alders and kill them 
At Muir Point by surface age 96 yea 
and Sandgren (1976) have recently s 
vegetation has developed with varie 
sisting of 126 recognizably distinc 
cular plants and lichens. Sprague 
already found in 1952 at surface ag 
abundance of parasitic and saprophy 
this site including a number of mus 
forming a fairy ring. 

since melting 
p to 10 m (33 
t (Fig. 2H,I) 
after ice re- 
rect, at some 
alders pro- 
nitrogen for 
of succession" 
prepare the 
bring about 
sy-turvy, for 
s rise above 
by shading . 
rs, as Noble 
hown, a dense 
d flora con- 
t taxa of vas- 
(1960) had 
e 72 years , an 
tic fungi at 
hrooms , one 

The later stages of development of vegetation 
and soil on Neoglacial till farther down the bay 
will not be dealt with here; they have been de- 
scribed by Cooper (1939) , Crocker and Major (1955) , 
Lawrence (1958) , Decker and Ugolini (1966) , and 


Reiners et al. (1971). Let us, instead, follow 
in detail changes at a given site over a period 
of 88 years following ice recession. 


When Cooper first visited Glacier Bay in 1916 
he established nine permanent plots, each one 
meter square, three on each of three surfaces of 
known age: 17, 24, and 37 years (Cooper 1923) 
since glacial recession where John Muir (1895) 
had sketched the positions of ice fronts in 1879, 
and Reid (1896) had mapped them in 1892, and 
Gilbert (1906) in 1899. One of these plots has 
been lost (destroyed by wave erosion before 1921) ; 
the other eight have been remapped and rephoto- 
graphed eight times between 1916 and 1972, four 
times by Cooper in 1916, 1921, 1929, and 1935, 
and the rest by myself, beginning in 1941. These 
permanent plots now have the longest history of 
observation in Alaska, and may be the oldest 
permanent plots in the world, on terrain of known 
age following glacial recession. 

Rationale for establishing the permanent plots. 
--In Cooper's words (1923: 355-365) "We are for- 
tunate, here at Glacier Bay in having a most un- 
usual opportunity to apply the method of the 
permanent quadrat, in the use of which we obtain 
accurate, unimpeachable data as to the movements 
and activities of the plant population of small 
areas typical to the whole." He emphasized "the 
unique advantage of this region for exact succes- 
sional study; the rapidity of vegetational devel- 
opment, and especially the known history of gla- 
cial behavior extending back a century and a 
quarter." Cooper was referring to Vancouver's 
visit and his map of the mouth of the bay in 1794 
(see Fig. 1) , when the terminal face of the ice 
had receded 10 km (6 mi) from its maximum ex- 
tent at the mouth of the bay, and subsequent 
maps by geologists and others showing the location 
of the ice margins in different years. He con- 
tinued: "It is thus possible, through a study 
of localities of known subaerial age, to deter- 
mine the net results of vegetational activity in 
a given period of years, and to make comparisons 
between areas of different lengths of history. 
Moreover, because of the rapidity of vegetational 
change, it is possible to watch the course of 
succession from year to year—actually to observe 
the development of the climax forest from pioneers 
to spruce and hemlock, almost complete within the 
bounds of a single lifetime." 

Enlargement of scope envisioned . --Cooper con- 
sidered the size of his plots adequate when he 
established them because of the small size of the 
plants. But he soon realized that they were too 
small. He wrote in 1923: "My study of permanent 
quadrats has been carried on, up to the present 
[1916 & 1921] upon a rather small scale, and yet 
results solidly grounded and of distinct value 
have already been obtained from it. It is my 
hope in the continuation of the investigation 
greatly to enlarge its scope." In his second 
trip in 1921 he actually enlarged one of the plots 
(#2) because the single Dryas plant had extended 
to fill the original square meter and far be- 
yond, expanding its cover "over 1000%" in five 
years. But in the two subsequent research trips 
of 1929 and 1935 he became more interested in the 
glacial history, and the contribution which could 
be made by a study of the distribution and iden- 
tification of the fossil stumps that occur in 
so many places along the shore north from the 
mouth of the bay to the head of Muir Inlet 
(Cooper 1937). Furthermore, he found that as the 
plant cover of these small plots increased in 
density of numbers, diversity of species, and 

complexity of stature, from a sparse, nearly two- 
dimensional film of mosses, or of scattered seed- 
ling willow herbs (Epilobium latifolium) , or al- 
pine avens ("Dryas drummondi i ) mats, or tufts of 
young scouring rush (Equisetum variegatum) , or 
creeping willow shrubs (Salix spp.;, the time re- 
quired to rechart and rephotograph the eight sur- 
viving plots (Plot #9 had been washed away by 
waves) , was all the time he could afford. In thos 
days, boats had to be chartered out of Juneau or 
beyond at increasingly higher daily rates. In sub 
sequent trips beginning in 1941 I have continued 
the study of these plots and have conducted a num- 
ber of other studies, but I have added only one 
plot, also a single square meter, in the isolation 
of Johns Hopkins Inlet near Tyeen Glacier 22 km 
(13 mi) southwest of Cooper Plot #1. 

Descr iption of the study si tes . --Cooper de- 
scribed the location of his permanent plots in 
great detail, because, as he pointed out: "Since 
my own activities in Glacier Bay sooner or later 
must cease, and since a study of such permanent 
areas grows continually more valuable with increas- 
ing time, it is advisable to put on record the 
exact position and means of location of each quad- 
rat, with the hope that some future visitor may 
carry on the work." I shall quote an example of 
his description for the area I discuss here in de- 
tail . 

"Station 26 (quadrats 1-2-3) is on the north- 
east shore of Reid Inlet [now considered the upper 
part of Glacier Bay proper] , opposite the center 
of Russell Island, and just southeast of a rather 
extensive alluvial fan. A short gravel spit, 
visible only at half-tide or lower, forms a small 
cove [a circular ice-block pit now called "Teacup 
Harbor"] protected from ice, with good anchorage, 
opening southwestward . ... Go ashore on north side 
of cove; go north a few paces to a conspicuous dar] 
glacially smoothed boulder 8 by 12 feet and 5 feet 
high, marked on opposite sides and top with white 
crosses." The paint was nearly invisible by 1941 
so I repainted and added a white marble capstone 
on my first visit that year (Fig. 3A) 

"Quadrat 1. From this boulder go north 37° 
east, magnetic bearing, 43 yards, to cairn 3 feet 
high, at northeast corner of quadrat. This is 
marked by stones at corners with white stripes 
bordered with black to show direction of bounding 
strips which run N. S. E. W. Station for oblique 
photograph is 6 feet from center of quadrat and 
marked by white cross on black ground." On a latei 
visit he installed steel rods as corner markers. 
These instructions have led me without difficulty 
to Plot #1 in 1941 and subsequent years, but re- 
locating most of the other plots would have become 
impossible if I had not at each visit photographer 
from special vantage points toward each plot and 
away from each plot. In 1972 a metal detector wa 
needed to relocate the iron corner rods at one pi" 
where they were completely submerged by Dryas mat 
In that year, with the help of Park Service Natur .1 
list Bruce Paige, I installed a pair of durable 
new markers near each of the plots with the plot 
midway between them to help insure the terms of 
Cooper's wish for their long continued relocation 
and study. 

History of Cooper Plot H1.--I shall trace here 
briefly the changes that we have observed since 
1916 in only this one plot, partly for lack of 
space and time, but mainly because it is the most 
accessible, the most easily relocated, one of the 
three on the oldest surface (now nearly a centur; 
and it lies on a nearly level surface (slope ang! 
12°) . Inspection of Figure 3 (top) shows the frc 
of the Grand Pacific Glacier extending northwest 


from Russell Island as it looked to John Muir in 
the autumn of 1879 when he first saw it. It was 
at the ice edge of that year that Cooper establish- 
ed his Plot #1 in 1916. By then the glacial front 
had receded 23 km (14 mi) up to Tarr Inlet to the 
Canadian border (Fig. 1). However, the develop- 
mental history of this plot's vegetation is atypi- 
cal in that we know of only a few seedling Dryas 
plants that have been seen on it, and these did 
not survive long. Cooper's 1929 chart (1931:74, 
Fig. 3) shows an area reputedly occupied by Dryas, 
but in a later article (1939: 147) he indicated 
that it was actually a new mat-forming moss, 
wrongly recorded on the published 1929 chart as 
"Dryas". In 1916, when first studied (Cooper 
1931: 74), it had about 30% cover of Rhacomi tr ium 
moss mats, 5 well-established prostrate shrub 
willows (4 Salix bar clay i , 1 S. arctica, 2 Carex 
sedges, nearly a hundred scouring rush tufts 
(Equisetum var iegatum) , about 34 willow seedlings 
(mostly S. barclayi) , and a tiny tuft of Stereo- 
caulon lichen. Changes over the next 19 years 
included increase in established willows to 14 
with gradual increase in height of stems, but 
still less than 0.3 m (1 ft) tall. Carex sedges 
increased in numbers and the scouring rush held 
its own, but over half of the plot area was still 
bare in 1935 (Cooper 1939: 142, Fig. 7). Progress 
from 1941 to 1967 can be seen at a glance in 
Figure 3 both for the area (Cooper's Station 26) 
as a whole (A, C, E) and within his Plot #1 (B, D, 
F) . Much bare gravel was still evident in 1941, 
62 years after glacial recession. Mo alder shrubs 
(Alnus crispa subsp. sinuata) were yet established 
either in 1941 or 1949, and the nearest fruiting 
alder that year was 37 m (120 ft) away and 2.4 m 
(8 ft) tall. The nearest tree species noted were 
a Sitka spruce (Picea sitchensis) 13.7 m (45 ft) 
away and 0.6 m (2 ft) tall, and a black cotton- 
wood ( Populus balsami fera subsp. tr ichocarpa) 
12.2 m (40 ft) away and 1.8 m (6 ft) tall. 'But 
by 1955 there was one alder established within 
the plot near the closer edge; it was in its third 
or fourth year of growth since germination, 0.53 m 
(21 in) tall, and its corrugated leaves are seen 
near the center of the photo (Fig. 3D) ; it had 
grown 0.43 m (17 in) taller in that season, 81% 
of its whole height. There were numerous alder 
seedlings but these all died later from shading 
by the rapid growth of this first established 
alder. A soapberry shrub (Shepherdia canadensis ) 
rooted outside the far side of the plot spread 
its branches over half the plot but its maximum 
attainable height is only about a meter here. 
Both these shrubs have nodulated roots, and their 
star mold symbionts fix atmospheric nitrogen, 
accounting for their rapid increase in size. 
Furthermore, the nitrogen in the amino acids and 
proteins in their fallen leaves incorporated in 
the soil and perhaps direct transport of nitro- 
gen compounds via mycorrhizal fungus threads have 
stimulated the formerly prostrate willows to send 
up erect shoots in that year 0.3 m (12 in). By 
surface age 88 years (Fig. 3E, F) the alder canopy 
over the plot was 4.6m (15 ft) tall and provided 
a 98% cover; all the willows and the soapberry had 
died because they were incapable of growing as 
fast and tall as the alder. All the Rhacomi tr ium 
moss was dead beneath the dense alder leaf litter. 
Numerous shinleaf herbs (Pyrola asar i folia) had 
become established and were thriving, some even 
flowering. The scouring rushes ( Equi setum varie- 
egatum) were thriving in spite of the dense shade 
and had taken on a deep bluegreen color, but bore 
no spore clusters (strobili) . A cottonwood tree 
crown similar to that in Fig. 3E, right of center, 
stimulated to rapid growth by the alder leaf 
litter had surpassed the alder canopy, and had 
developed a trunk 7.6 cm (3 in) in diameter; it 

stood only 3.2 m (10.5 ft) outside the plot and 
must have been established as a seedling many 
years before and could have persisted unnoticed 
as a yellowish sickly nitrogen-starved prostrate 
individual for over 40 years. At the time of my 
most recent visit to the plot in August 1972 at 
surface age 93 years, the number of shinleaf plants 
had decreased markedly since 1968, probably due to 
shading and the alder leaf fall, which had accumu- 
lated a rich brown organic layer 8.9 cm (3.5 in) 
deep on top of the pebble pavement of the mineral 
substratum. The larger fragments of alder twig 
litter and the bases of the living alder stems 
were covered with moss. These mosses, the few 
remaining shinleaf plants, and the numerous wispy 
blue-green stems of the scouring rush, still with- 
out spore clusters, formed the only visible green 
below the dense alder canopy. There were many tiny 
fleshy fruiting bodies of ascomycetous fungi di- 
gesting the fallen alder twigs and leaves. Since 
it was no longer practicable to photograph the 
plots from above, I began in 1967 to photograph 
upward from the plot center with a fisheye lens 
camera to record changes in canopy cover. 


Lituya Bay and its Neoglacial and giant wave 
his tor y . --This T-shaped bay (Fig. 4) was "dis- 
covered", mapped, and named "Port des Francais" by 
La Perouse in July 1786 (1799). Indian fishing 
villages were then on the shores near the mouth of 
the bay. He and his crews in two vessels stayed 
there about a month. His excellent map of the 
whole bay is reproduced in Post and Streveler 

(1976: 113, Fig. 2). It shows that the termini of 
Lituya and North Crillon glaciers which lie at the 
heads of the bay in the transverse trench of the 
Fairweather Fault, had receded then farther than 
their present positions by about 6 km (4 mi) and 
2.8 km (1 3/4 mi). Thus recession of Neoglacial 
ice began earlier than at Glacier Bay, probably by 
as much as 200 years, or about A.D. 1550 (Mark 
Noble, personal communication 1976, based on tree 
cores) . The general glacial history has been 
worked out by Goldthwait, McKellar and Cronk 

(1963). The main reach of the bay extends north- 
east about 12.1 km (7 1/2 mi) from the mouth, and 
contains a single central island named Cenotaph 
by La Perouse. The western slope of the Fair- 
weather Range rises northeastward abruptly within 
3 . 2 km (2 mi) of the fault trench to elevations of 
over 1829 m (6000 ft) . This steep bare bedrock 
slope is heavily fractured and it is easy to visu- 
alize how parts of it could break off, especially 
at times of earthquake, and roar into the bay, 
setting up giant waves. Fortunately, the existence 
of such waves had been surmised, and a careful 
study of their effects had been begun by Don J. 
Miller in 1952, and dates of earlier giant waves 
had been worked out in 1952-53 by studying tree 
cross-sections from along the wave trimlines. The 
physiographic accidents occurred in 1853 or 54, 
about 1874, possibly in 1899, and in 1936 perhaps 
by some different mechanism. The most recent wave 
was on July 9, 1958, induced by an earthquake with 
horizontal slipping along the Fairweather Fault 

(Goldthwait personal communication 1973). This 
history has been beautifully documented and pre- 
sented by Miller (1960) . Of the three fishing 
boats in the bay at the time, the occupants of two 
survived to provide eyewitness accounts. The most 
recent wave seems to have been larger than the 
others, for it demolished the forest up to and 
beyond most of the earlier wave trimlines; only a 
portion of the forest grown up since the old 1854 
wave survives as a narrow strip on the mainland 


FIGURE 3. Vegetational development, primary succession on glacial till over an 88-year span at Cooper 
Station 26 (top photo, and A, C, E) , and on his nearby Point #1 (B. D. F) following recession of Grand 
Pacific Glacier. 


northeast of Cenotaph Island (Fig. 4, and Fig. 
5B at left of center) . The more numerous decid- 
uous tree crowns of black cottonwood distinguish 
it from the older more exclusively conifer forest 
above. The cliff face from which the rockslide 
broke out is marked on the map (Fig. 4) by the 
stippled area, and in Figure 5, top photo, copied 
from Miller (1960), by the letter "r", and in the 
Figure 5A where the M-shaped snow area appro- 
priately commemorates the name of Don J. Miller 
who first correctly identified giant waves as the 
cause of the forest destruction. The 1958 rock- 
slide started from an elevation of about 914 m 
(3000 ft), struck the lower part of Lituya Glacier, 
sheared off about 400 m (1300 ft) of its tip, 
and splashed up the opposite wall of the fault 
trench valley to an elevation of 524 m (1720 ft) ! 
Based on measured elevations along the forest 
trimline the water wave ascended with destructive 
force as much as 207 m (680 ft) on the south 
valley wall near the head of the main bay and 
then its crest gradually diminished down the bay, 
but it was still able to destroy the forest up 
to an elevation of 9.14 to 10.67 m (30-35 ft) 
near the mouth of the bay (Miller 1960: 58, Fig. 
15) . In the lower relatively level areas the 
salt water invaded the forest without destroying 
it but inflicted salt damage that became visible 
later that growing season. 

Eyewitness accounts . — The captains of both 
fishing boats that survived first noticed violent 
vibrations about quarter past 10 local time 
(about sunset) . Mr. and Mrs. Swanson had anchored 
their 40-ft boat, the Badger in a cove of the bay 
just north of the entrance. About a minute later 
Swanson looked toward the head of the bay past 
the north end of Cenotaph Island and saw that 
Lituya Glacier had risen up in the air, jumping 
and shaking, and had moved forward into view. 
Then the glacier dropped down out of sight and a 
big wall of water came over the north point at 
the head of the bay, crossed over and swept high 
up the opposite shore, then sped down the bay 
past Cenotaph Island. The wave reached the 
Badger, still at anchor, about 4 minutes after it 
was first sighted. The Badger was lifted by the 
wave and carried like a surf board, stern first, 
across the north spit. Swanson looked down on 
the trees growing on the spit from a height he 
estimated to be about 24 m (80 ft) above the tree 
tops. The wave crest broke just outside the spit 
and the boat hit bottom and foundered some dis- 
tance from shore. The Swansons abandoned their 
boat in a small skiff and were picked up by 
another fishing boat about two hours later. The 
narrator of the other account, Mr. Ulrich, and 
his son, in their boat Edrie, were anchored in 
the bay near the south shore about 2 km (1.2 mi) 
from the mouth. He promptly started his engine 
when he felt the earthquake vibrations, and about 
2 1/2 minutes later he heard a deafening crash 
from the head of the bay. The anchor chain 
snapped as the wave 15-23 m (50-75 ft) high 
approached, but the boat rode it out. After the 
giant wave passed the water returned to normal 
level but was very turbulent for about a half 
hour with much sloshing about from shore to shore 
with steep sharp waves up to 6 m (20 ft) high. 
This perhaps accounts for the parallel pair of 
wave-rows of logs along the denuded slopes north- 
ward from Cenotaph Island as noted by Miller 
(1960: 58, Fig. 15). Within an hour of the 
passage of the giant wave Ulrich noted that the 
bay near shore was filled with floating logs, but 
the Edrie passed safely out of the entrance of 
the bay by about 11:00 p.m. 

The secondary succession. — We may define secon- 
dary succession as the re-development of eco- 
systems following occurrences which destroy the 
aerial portions but not all the soil and its 
living contents. Unfortunately, the opportunity 
which this geomorphic event provided for study of 
secondary succession was not taken advantage of 
until several years had passed. Streveler had 
made some preliminary observations (personal com- 
munication) . But our 1973 visit of June 19-25 on 
the Park Service motor vessel Nunatak with Captain 
James Sanders, was an initial attempt to collect 
some more comprehensive information. We are 
grateful to Park Service Superintendent Robert E. 
Howe for providing the necessary transportation. 
Field assistance was provided by Professor Edward 
J. Cushing. Our major objective was to observe 
rates of vegetational redevelopment and the numbers 
and diversity of the plants involved in the secon- 
dary succession here for comparison with those of 
the primary succession on deglaciated surfaces 
already studied as mentioned above at nearby 
Glacier Bay. We selected an area within the 1958 
wave-denuded belt along the south mainland shore 
directly south of Cenotaph Island (circled dot in 
Fig. 4, and Fig. 5 top photo) because it was con- 
venient to the anchorage Captain Sanders had se- 
lected for the Nunatak. Here we found a denser 
stand of woody vegetation grown up within 15 years 
after the 1958 giant wave destruction of the 
previous forest cover than had developed at 
Glacier Bay on a surface of raw glacial till in 
50 to 100 years. Not only did we study the new 
vegetation on the partly wave-eroded glacial till, 
but we examined the heaps of logs that had been 
assembled in giant piles and wave-rows as seen in 
Figure 5C and E. On the logs close to the bay 
shore there were only lichens, but on the south 
slope of the bay, near our study plot where the 
major wave-row of logs was piled 4.6 m (15 ft) 
deep against the wave trimline, young hemlock 
trees, wnich favor logs for seedling establishment, 
were already about 2 m (7 ft) tall. They were 
rooted on the crest of the wave-row, perched high 
above the mineral soil and presumably will be 
gradually lowered to the soil surface as the logs 

The permanent plot. — A large circular permanent 
plot of 5.18 m (17 ft) radius and 82.32 sq m (751 
sq ft) was established on a Level terrace of the 
mainland south shore (Fig. 4 and Fig. 5 top photo) 
the data presented below are given in terms of an 
area of 100 m 2 (1076 ft 2 ) . The plot is centered on 
a marked dominant red alder tree (Alnus oregona) . 
It is situated about 400 m (1/4 mi) northeast of a 
pond (Fig. 4) . This tree was estimated to be about 
11.9 m (39 ft) tall, 14.6 cm (5 3/4 in) dbh (dia- 
meter at breast height) and about 12 years old. It 
is marked by a 1.32 m (52 in) length of galvanized 
steel tubing hung on the magnetic west side of the 
trunk with a loop of insulated copper wire coiled 
to allow for trunk growth. Three boulders, the 
largest 91 cm (3 ft) long, lie in a cluster at the 
west base of this tree. A marker tube 38 cm (15 
in) long was driven into the soil at the magnetic 
S end of this largest boulder so that its top pro- 
truded 14.8 cm (5 7/8 in) above the mineral soil. 
Another marker tube 1.32 m long lies horizontally 
southward on the forest floor attached at one end 
to this driven marker. Both the hanging and the 
horizontal tubes were appropriately labeled in 
soft pencil, a type of label which, in Minnesota, 
is still legible after 20 years. 

The earliest stage, beginning with willows and 
black cottonwoods seems to have been similar to 
that at Glacier Bay, but the date of the giant 
wave, July 9th, 1958 seems to have been too late 
in the season for dispersal and establishment of 



FIGURE 4. Map of Lituya Bay which had been nearly emptied of glacier ice by 1786 when La Perouse 
visited and mapped it. Its shores have been partly denuded by a series of giant waves, the most recent 
on July 9, 1958, initiated by earthquake and rockslide (strippled) falling 914 m (3000 ft). The wave 
splashed up to a maximum evelation of 524 m (1720 ft) and swept away forest but not all soil along the 
lower banks as may be seen in the top photo of Figure 5 taken a month after the event. Subsequent second- 
ary succession on the denuded surfaces has been extremely rapid and vigorous, dominated on the south 
shore by tree alders grown 12 m (40 ft) tall in 15 yrs. Neoglacial maximum here was 200 years earlier 
than in Glacier Bay, but some of that ice is still melting from beneath a tilting spruce and hemlock 
forest north of Fish Lake. 

these species that first year because their seeds 
ripen early and are very short-lived, but seeds 
of alders, spruce, and hemlock ripen late, so 
may have entered in the autumn of the same year. 
We have seen no evidence that Dryas has been in- 
volved at all at Lituya Bay. Although the seeds 
of both Sitka alder (Alnus crispa subsp. sinuata) 
a shrub, and red alder (Alnus oregor.a) , a tree, 
very likely entered the plot area in autumn of 
1958, they could not have germinated until the 
following spring, 1959, because they needed a long 
cold moist period to break dormancy. Densities 
and heights of these various woody plants reveal 
important facts about the developing community, 
but it is unfortunate thr.t some similar data were 
not collected a decade earlier. The red alder 
trees occurred in the plot at a density of 32 
trunks per 100 m*, and were 8.25 cm (3.25 in) 
average dbh, and up to 12 m (40 ft) tall. Tho 
capacity of this symbiotic nitrogen-fixing tree 
for height growth beyond that of any of its asso- 
ciates has resulted in deep shade below its 
canopy, with weakening and death of the shorter 
shrubs including the Sitka alder which occurred 
at a density of 8 shrubs per 100 m^ with 2-4 
per cluster, 3-5 m (10-16 ft) height and 1.5-4 
cm (0.6-1.57 in)dbh. The remains of the dying 

and dead shrubs were still present (Fig. 5F) ; the 
only thrifty Sitka alder was in a slight canopy 
opening nearby. There were only 5 black cotton- 
woods per 100 m , averaging 3.8 m (12.5 ft) tall 
and 2 . 5 cm (1 in) dbh; most had been severely de- 
barked by porcupines above the 1.5 m (5 ft) level 
presumably above the snow blanket during the pre- 
vious two winters and only one seemed healthy 
enough to survive. But beneath the dense shade 
of the tree alders was a thriving population of 
young evergreen conifers (Fig. 5F , a photo 100 m 
east of the plot) . There Were Sitka spruces 
(Picea si tchensi s) 85 per 100 m , averaging 65 
cm (2.14 ft) tall and 9.7 years old (range 6-14 
yrs) based on branch whorl counts. Western hem- 
locks occurred at a density of 18 per 100 m^ 
averaging 37 cm (1.12 ft) tall. Associated with 
these most abundant woody plants were 5 other 
shrubs, 12 herbaceous seed plants and 5 mosses. 
Two of the plant species adjacent to the plot, 
devils club shrub (Echinopanax horridum) and 
western skunk-cabbage (Lysochiton americanum) 
occur here though I have never seen them on the 
deglaciated Neoglacial terrain at Glacier Bay. 
I believe they survived here through forest 
destruction and brief immersion in salt water as 
underground roots and stems which regenerated 
sprouts . 


FIGURE 5. Scenes at Lituya Bay. Top, 1958, a month after the giant wave; others late June 1973, 15 yrs 
later. A-C on left side; D-F on right side. C and E show forest wreckage, F new red alder forest with 
trees 12 m (40 ft) tall, with dead Sitka alder shrub in foreground and thriving understory of Sitka spruce 
and western hemlock, 100 m east of Plot #73-1. Top photo Miller (1960). 

This rapid development of the community, with 
accumulation of 4.5 cm (1 3/4 in) of organic 
matter, mostly alder leaf litter, on the soil but 
not on the boulders, has come about mainly be- 
cause of the close proximity of Negolacial and 
pre-Wisconsin refugium forests containing the 
red alder, Sitka spruce and western hemlock. 
The accumulation of great biomass made possible 
by nitrogen fixed by the alder root nodule sym- 
bionts and inherent potential for attaining tree 
stature seems to be very large compared to that 
of the comparable stages in Sitka alder shrub 
communities 50 to 100 years following glacial 
recession at Glacier Bay just 60 km (37 mi) to the 
east. Measurements of biomass have not been done 
either at Glacier Bay or at Lituya Bay. Such 
studies would provide important contributions to 

our knowledge of the rate of primary and secondary 
ecosystem development in this maritime subarctic 


BECKING, J. H. 1970. Plant-endophy te symbiosis 

in non-leguminous plants. Plant and Soil 

BIERHORST, D. W. 1971a. Morphology of vascular 

plants. New York: Macmillan. 
BIERHORST, D. W. 1971b. Morphology and anatomy 

of new species of Schizaea and Actinostachys . 

Amer. J. Bot. 58:634-648. 
BOHN, D. 1967. Glacier Bay: the land and the 

silence. San Francisco: Sierra Club. 

165 p. 


COOKE, W. B. and D. B. LAWRENCE. 1959. Soil 
mould fungi isolated from recently glaciated 
soils in south-eastern Alaska. J. Ecol . 
COOPER, W. S. 1923. The recent ecological 
history of Glacier Bay, Alaska. Ecology 4: 
93-128, 223-246, 355-365. 
COOPER, W. S. 1931. A third expedition to 

Glacier Bay, Alaska. Ecology 12:61-95. 
COOPER, W. S. 1937. The problem of Glacier 
Bay, Alaska: A study of glacier variations. 
Geog. Rev. 27:37-62. 
COOPER, W. S. 1939. A fourth expedition to 
Glacier Bay, Alaska. Ecology 20:130-155. 
COOPER, W. S. 1942. An isolated colony of 
plants on a glacier-clad mountain. Bull. 
Torrey Bot. Club 69:429-433. 
COOPER, W. S. 1956. A contribution to the 

history of the Glacier Bay National Monument. 
Minneapolis: Unviersity Minnesota, Dept . 
Botany. 36 p. 
CROCKER, R. L. and J. MAJOR. 1955. Soil develop- 
ment in relation to vegetation and surface 
age at Glacier Bay, Alaska. J. Ecol. 43:427- 
DECKER, H. F. 1966. Pages 73-95 in: Goldthwait 

et al. (1966) . 
FIELD, W. 0. 1947. Glacier recession in Muir 

Inlet, Glacier Bay, Alaska. Geog. Rev. 37:369- 
FIELD, W. O., editor. 1975. Mountain glaciers 
of the Northern Hemisphere. 2 vols + atlas. 
Hanover, NH: US Army Corps of Engineers. Cold 
Regions Research and Engineering Laboratory. 
GILBERT, G. K. 1904. Glacier Bay. Pages 16-39 
in: Glaciers and glaciation. Vol. Ill of 
Alaska. Harriman Alaska Expedition. New 
York: Doubleday, Page. 4 vols. 
(A. Mirsky, editor) 1966. 
and ecological succession in a deglaciated 
area of Muir Inlet, southeast Alaska. Ohio 
State Univsty Inst, of Polar Studies, Report 
No. 20. 167 p. GOLDTHWAIT, R. P., I. C. MCKEL- 
LAR, and C. CRONK. 1963. Fluctuations of 
Crillon Glacier system, southeast Alaska. 
Ohio State Univ. Inst. Polar Studies, Contrib. 
No. 35. 11 p + 5 figs, (mimeo) . Originally 
publ. in IUGG Internat. Assoc. Sci. Hydrol. 
Bull. 8, No. 2, pp 62-74. 
GRIGGS, R. F. 1933. The colonization of the Kat- 
mai ash, a new and inorganic "soil." Amer. 
J. Bot. 20:92-113. 
GRIGGS, R. F. and D. READY. 1934. Growth of 
liverworts from Katmai in nitrogen-free 
media. Amer. J. Bot. 21:265-277. 
INGOLD, R. S. and A. T. NAVARRE. 1952. "Pollut- 
ed" water from the leaching of igneous rock. 
Science 116:595-596. 
LA PEROUSE, J. F. A. 1799. A voyge 'round the 
world 1785-1788. Transl. from the French. 
London: 3 vols and folio. 


Soil development 

LAWRENCE, D. B. 1958. Glaciers and vegetation 
in southeastern Alaska. Amer. Scientist 46: 

G. BOND. 1967. The role of Dryas drummondii 
in vegetation development following ice reces- 
sion at Glacier Bay, Alaska, with special 
reference to its nitrogen fixation by root 
nodules. J. Ecol. 55:793-813. 

MILLER, D. J. 1960. Giant waves in Lituya Bay, 
Alaska. USGS Prof. Paper 354-C. Washington, 
D. C: US Govt. Ptg. Off. pp i-iii, 51-86, 
7 figs. , 9 plates . 

MUIR, J. 1895. The discovery of Glacier Bay. 
The Century Magazine 50, N.S. 28:234-247. 

NOBLE, M. , and C. D. SANDGREN. 1976. A floris- 
tic survey of Muir Point, Glacier Bay Nation- 
al Monument, Alaska. Bull. Torrey Bot. Club 

POST, A. and G. STREVELER. 1976. The tilted 

forest: glaciological-geologic implications of 
vegetated Neoglacial ice at Lituya Bay, Alas- 
ka. Quaternary Research 6:111-117. 

REID, H. F. 1896. Glacier Bay and its glaciers. 
USGS 16th Annual Rept. 1894-1895. 

1971. Plant diversity in a chronosequence at 
Glacier Bay, Alaska. Ecology 52:55-69. 

RIDLEY, H. N. 1930. The dispersal of plants 

throughout the world. Ashford, Kent, England: 
L. Reeve. 

SCHOENIKE, R. E. 1958. Influence of mountain 
avens (Dryas drummondi i ) on growth of young 
cottonwoods (Populus trichocarpa) at Glacier 
Bay, Alaska. Minnesota Acad. Sci. Proc. 

SINGH, R. N. 1960. Role of blue-green algae in 
nitrogen economy of Indian agriculture. New 
Delhi: Indian Council Agric. Res. 

SPRAGUE, R., and D. B. LAWRENCE. 1960. The fungi 
on deglaciated Alaskan terrain of known age. 
(Part III) Washington State Univ. (Pullman) 
Res. Studies 28(l):l-20. 

STEWART, W. D. P. 1966. Nitrogen fixation in 
plants. London: Athlone. 

STREVELER, G. P., and B. B. PAIGE. 1971. The 
history of Glacier Bay National Monument, 
Alaska: a survey of past research and sugges- 
tions for the future. USDI National Park 
Service, Glacier Bay National Monument. 89 p. 

UGOLINI, F. C. 1966. Soils. Pages 29-58 In: 
Goldthwait et al. (1966). 

(informational folder with map in color) . 

WATTS, W. A. 1967. Late-glacial plant macro- 
fossils from Minnesota. Pages 89-97 in: 
Quaternary Paleoecology . Edited by H. W. 
Wright and E. J. Cushing. New Haven, CT: 
Yale University. 

WORLEY, I. A. 1973. The "Black Crust" phenome- 
non in upper Glacier Bay, Alaska. Northwest 
Science 47:20-29. 


John Ewel and Louis Conde 


The Everglades National Park recently completed 
purchase of a large block of private holdings 
surrounded by Park lands. About 4000 ha of the 
purchased lands had been farmed, but will now re- 
vert to natural vegetation. Park managers are 
faced with a variety of questions: What kinds of 
ecosystems will appear on these lands? Will ex- 
otic plants be important, conspicuous components 
of the flora? Are seeds from the surrounding ma- 
ture ecosystems getting into the abandoned farm- 
lands? Do the species found on the farmlands pose 
a threat to the native, mature ecosystems which 
surround them? Can the new vegetation be manipu- 
lated by controlling the presence of seed sources, 
either through addition or extermination? Partial 
answers to some of these questions require knowl- 
edge of the kinds and amounts of seed present in 
the various successional and mature ecosystems. 
This study reports our findings of an inventory 
of the seed content of surface soils of five 
successional and two mature Everglades Park eco- 
systems. Our work was supported, in part, by 
U.S.F.S., Southeast Forest Exp. Station, Research 
Agreement 18-492. 

An excellent review of the floristic role of 
seeds in the soil, as well as the problems in- 
volved in assaying the viable seed content of 
soil, was compiled by Major & Pyott (1966) . They 
describe four main problems related to such 
studies: aggregation of propagules near mother 
plants; vegetative reproduction; sampling un- 
knowns (depth, area, and number) , and selective 
germination caused by the particular conditions 
to which the samples are subjected. In our study, 
conditions (well-watered flats exposed to full 
sun) clearly favored germination of weedy species 
and those which did not require partial shade or 
special pretreatment such as passage through an 
animal's digestive tract. 


The recently purchased farmlands are collec- 
tively referred to as the Hole-in- the-Donut be- 
cause they constitute an island of former farm- 
land completely surrounded by mature plant 
communities. Some parts of the Hole-in-the-Donut 
have been unfarmed for about 35 years, while 
others were farmed as recently as June 1975; most 
of the land has been abandoned for less than 
five years. About 90 percent of the Hole-in-the- 
Donut had been wet prairie or sawgrass glade 
prior to farming, while most of the remainder was 
pineland (Hilsenbeck 1976) . Fire is important in 
the maintenance of both the prairie-sawgrass 
communities and the pinelands. The vegetation, 
soils, hydrology, and role of fire in these eco- 
systems have been described by Harshberger (1914), 
Davis (1943), Egler (1952), Robertson (1953), 
Craighead (1971), and Alexander & Crook (1973). 
The wet prairie and sawgrass {cladium jamaicensis , 
a sedge) communities occupy marl soils which 
cover the limestone bedrock to a depth of about 
10 cm. These herbaceous communities are period- 

Botany Department, University of Florida, 
Gainesville 32611 

ically flooded between June and November. The 
pine community (pinus elliottii var. densa) 
occupies slightly higher ground along the north- 
ern rim of the Hole-in-the-Donut; the shallow 
soils which cover the limestone there are loam 
rather than fine-textured marl and are not 
usually flooded during the wet season. 

Everglades ecosystems are prepared for winter 
vegetable farming by rock plowing. This process 
breaks up the limestone to a depth of about 20 
cm and the crushed limestone is mixed with the 
overlying marl; a bedding plow is then used to 
create alternating beds and furrows. Reclamation 
efforts by park managers have concentrated on 
mowing the successional vegetation, followed by 
twice disking the land to eliminate the beds and 
furrows. Various combinations of mowing and fire 
are being studied as possible tools to keep the 
vegetation in an early successional stage until 
more is learned about its probable course of 
succession. Without fire or mowing it appears 
that rock-plowed prairie and sawgrass communities 
develop into forests after farming stops. These 
new forests contain numerous tree species, but 
are often dominated by Schinus terebinth! folius , 
an exotic tree considered by most to be an un- 
desirable component of the Park flora. The 
second-growth vegetation in the Hole-in-the-Donut 
and its possible management by various combina- 
tions of mowing and burning are currently the 
subject of a detailed study being conducted by 
Park personnel (see Hilsenbeck 1976 and Res. 
Mgt. Staff 1976) . 

The seven study sites were all located in, or 
contiguous to, the Hole-in-the-Donut. The sample 
areas were all within 7 km of one another and each 
site except the mature forest had at least one 
border in common with the most recently abandoned 
farmlands. The five successional stages of 
vegetation sampled were: 

(i) New field: This site was formerly glade 
vegetation and is flooded from June 
through November. It had been farmed 
until June 1975 and was disked in November 
1975. The site is contiguous to both the 
mature glade and the 35-year-old forest. 

(ii) 2.5-year-old, burned in June 1975: The 
original vegetation was glade, near its 
boundary with pineland. The vegetation 
on the site was dominated by Ludwigia 
erecta and L. octovalvis, which are woody 
shrubs about 2 m tall. The burn (part 
of a Park experiment) was spotty and in- 
complete. The site is located near the 
middle of the Hole-in-the-Donut, 2.5 km 
from the mature glade and 4 km from site 
(iii) . 

(iii) 2 . 5-year-old , unturned : The original 
vegetation on the site was glade. The 
vegetation at the time of soil sampling 
was similar to that of the burned site; 
it was dominated by the same Ludwigia 
spp. plus Panicum bartowense . The site 


is located about 3 km from the 17-year- 
old woodland and 4 km from the burned site. 

(iv) 17-year-old woodland : This site was the 
only one located on former pinelands and 
was not subject to wet-season flooding. 
The dominant plant was Myrica cerifera 
(a native shrub about 5 m tall) but Schinus 
was also found. The site is located in 
the northwest part of the Hole-in-the-Donut 
about 3 km from the unburned site and 2.5 
km from the burned site. 

(v) 3 5-year-old forest: This site occupies 
former glade and prairie. The forest is 
dominated by the exotic species Schinus 
terebinthifolius and is the oldest 
successional vegetation available within 
the Hole-in-the-Donut. Beds and furrows 
are still evident. The site is contiguous 
to both the new field and the mature glade. 

Two mature ecosystems adjacent to the succes- 
sional vegetation of the Hole-in-the-Donut were 
also studied: the first because it represents the 
original vegetation for most of the study sites, 
and the second because the successional vegetation 
seems to be developing into a forest rather than 
an herbaceous community. 

(vi) Mature glade: This community consists of 
a diverse mixture of fire-adapted sedges, 
grasses, and f orbs , and is flooded from 
June through November. The site is 
contiguous to both the new field and the 
35-year-old forest. 

(vii) Mature forest: This complex assemblage 
of more than twenty woody species is 
typical of the subtropical hammock forests 
of south Florida. It is rarely subjected 
to flooding, and fire is not a normal 
environmental factor involved in its 
maintenance. It is located at the north- 
east corner of the Hole-in-the-Donut about 
1.5 km from the new field and the 35-year- 
old forest. 

On December 14, 1975 the surface soil (c. 3 cm) 
was removed from three 50 cm x 50 cm plots, 
located at 20 m intervals along a transect, at 
each site. The soil was transported to Gaines- 
ville, Florida, where it was stored in covered 
containers at room temperature until December 24 
when each soil sample was spread over a flat of 
sterilized greenhouse potting sand. The twenty- 
one flats were assigned to randomly selected 
locations on an unshaded greenhouse bench and 
watered daily. The flats were monitored for one 
year, during which seedlings were counted and 
harvested at frequent intervals to avoid mortality 
from crowding. Sample seedlings of each species 
were transplanted and grown to reproductive stages 
for identification (except for 19 species, repre- 
senting <1 percent of the individuals, which have 
not yet flowered) . Identifications were provided 
by D. Hall, FLAS Herbarium, University of Florida. 


The abundance and site distribution of the 
species which appeared in the flats are summarized 
in Table 1. The total sample area of 5.25 m 2 
yielded approximately 120 species and nearly 
33,000 individuals. We suspect, but cannot prove, 
that four of the species (Gnaphal ium pensylvanicum , 
Oxalis filipes, Pteris vittata, and Conyza 
canadensis) are greenhouse contaminants: all ex- 
cept p. vittata appeared in the flats from all 

sites, all are common weeds in and around our 
greenhouse, and none are conspicuous components 
of the Hole-in-the-Donut flora. Table 1 does 
not include a breakdown of the data among the 
three replicates per site, but in general, rep- 
licate samples were f loristically and numerically 
similar, whereas among-site differences were 
readily apparent, even upon casual observation of 
the flats . 

The 30,000 individuals were very unequally 
distributed among the 119 species (Fig. 1) . Most 
species were represented by ten or fewer indi- 
viduals, and more than 90 percent of the species 
were represented by less than 100 individuals. 
Some of the seventeen species represented by more 
than 100 individuals are typical invaders of old 
fields; they are prolific seed producers and 
widespread, but are short-lived and unlikely to 
permanently occupy the former farmlands unless 
these are subjected to further drastic soil 
disturbance. Included in this category are 
genera such as Amaranthus , Ambrosia , and Polygonum 
Their great abundance is typical of newly 
abandoned farmlands, and they are likely to be- 
come less important as succession proceeds. Some 
of the other species which were extremely 
abundant, however, are more long-lived and are 
likely to become semi-permanent components of the 
successional vegetation, especially if management 
practices such as mowing and burning arrest 
succession. Representative of this category are 
Baccharis , Ludwigia , Myrica, and various grasses, 
including Panicum bar towense . These plants are 
almost certain to be important components of the 
successional vegetation in the Hole-in-the-Donut 
and merit special attention from Park managers. 


The numbers of seeds varied greatly among 
sites (Fig. 2) , ranging from thousands per m 2 
in the younger successional stages to about 500 
per m 2 in the oldest successional forest and the 
two mature ecosystems. The 2.5-year-old stand 
which had been burned six months prior to sampling 
contained the most seeds: more than 30,000 per 
m 2 . Of these, 84 percent were Ludwigia octovalis 
+ L. erecta (two species which are inseparable in 
the early stages of growth) . Dominance of a seed 
flora by very few species is a commonly reported 
phenomenon (e.g., Kropac 1966, Jansen 1969, 
Kellman 1970, 1974a) . 

The numbers of 
soils from our yo 
(i.e. , c . 4500 to 
those reported in 
soils, pastures, 
sion (Kropac 1966 
1971, Kellman 197 
Higher seed densi 
literature, such 
samples which yie 
but these are usu 
deeper soil sampl 
Everglades. Jens 
to a depth of 20 

germinated seeds ob 
ungest successional 
■30,000 per m 2 ) are 
the literature for 
and very early stage 
, Jensen 1969, Feast 
4b, Lockett & Robert 
ties have been repor 
as one of Jensen's ( 
lded 66,800 viable s 
ally derived from a 
e than the 3 cm we u 
en (1969), for examp 
cm . 

served in 

similar to 
s of succes- 

6. Roberts 
s 1975) . 
ted in the 

eeds per m 2 , 
substantiall j 
sed in the 
le, sampled 

Likewise, the lower densities of germinated 
seed (usually less than 500 per m 2 ) observed in 
our oldest successional stand and in the two 
mature ecosystems are consistent with values re- 
ported in the literature for old secondary and 
mature forests (Olmsted & Curtis 1947, Keay 1960 
Kellman 1970, Guevara & Gomez-Pompa 1972, Liew 
1973, Kellman 1974a, Strickler & Edgerton 1976). 
The drop in numbers of germinated seeds per unit 
area from the early stages of succession of the 























1 8 

























3 5 




TABLE 1. Numbers of seedlings harvested from soil samples taken from seven Everglades ecosystems. 
Each value is a total of three 0.25 mr samples. Origin: I = indigenous; X = exotic. Weediness: 
W = weedy or occurring on disturbed sites (Long & Lakela 1971); N = non-weedy according to Long & 

Species A B C D E F G 

Acrastichum aureum L. 

Alternant her a romos iss i ma 
(Mart.) Chodat 

Amaranthus hybridus 

Ambrosia artemisiifolia L. 

Ammani a cocci nea Rottb. 

Ammania teres Raf. I N 165 28 

Aristida purpurascens Poir. 

Ascyrum hypericoides L. 

/Aster bracei Britt. IN 1 

Aster dumos us L. IN 

Azolla caroliniana Willd. 

Baccharis halimi folia L. N 23 47 67 

Bacopa monnieri (L.) 
Wettst . 

Bidens pilosa L. 

Bo ehmer ia drummondiana 

Borreria laevis (Lam.) I N 434 225 

Gr iseb . 

Borreria ocimoides IN 2 11 

(Burm. f) DC. 

Centella asiatica (L.) I W 2 


Chamaesyce hypericifolia I W 1 1 1 44 4 16 

(L. ) Mi lisp. 

Chamaesyce opthalmica I W 2 4 2 26 1 11 

C Per s . ) Burch 

Chenopodium album L. I W 7 1 1 

Cissus S icyoides L. I N 1 

Colocasia esculentum (L.) X W 2 

Commelina diffusa Burm. f. I N 1 11 

Conyza bonariensis (L.) X W 1 

Cronq . 

'Conyza canadensis (L.) I W 6 6 8 24 8 3 

Cronq . 

Conyza parva Cronq. I W 8 1 

Cus cuta campestris Yuncker IN 1 

Cynanchum sp. 1 

Cynoctonum mitreola (L.) I N 8 4 

Br i tt . 



Cynoctonum sess i li fol i urn I N 

(Walt.) J. F. Gmel. 

Cyperus escul entus L. 

Cy per us iria L. 

Cyperus ochraceus Vahl. 

Cyperus po lys tachyos Rottb. 

Cyperus sur i namens i s Rottb. 

Cyperus spp. (max. 7 spp.) 

Di chromena colorata (L.) I N 


Digitaria adscendens (HBK.) X W 

Henr . 


Eclipta alba (L.) Hassk. 

Eleocharis geniculata (L.) 
R. & S . 




















4 1 





Eleusine indica (L.) X W 

Gaertn . 

Eragrostis el li ott ii Wats. 

Erechtites hieraci folia 
(L.) Raf . 

Erigeron quercifolius Lam. 

Erigeron vernus (L.) T. & G. 

Er ys imum chei rant ho ides L. 

. a sp . 

Eupator ium capi 11 i fol ium I W 

( Lam . ) Small 

Eupator i um coel est i num L. IN 

Eupatorium leptophyllum DC. I W 

Evolvulus ser i ceus Sw . I N 

Ficus sp. 

Galium obtusum Bigel. 

■ :i um carolinianum L. 

Gnaphalium pensy lvanicum 
Wii: d . 

Gnaphalium spicatum Lam. 

Heliotropium polyphylluw. 
Lehm . 

Hydrocotyle verticillata I N 

Thunb . 

Hypoxis juncea Smith I N 

Hyptis alata (Raf.) I N 

Shi nners 

Kos te let zkya virginica I N 

(L. ) Presl ex Gray 


I W 1 15 3 
I N 3 1 1 








l l . 

4 ( < 






r ,i 

4 3 



4 138 13 5 

I N 

I W 1 1 3 7 

I W 
I N 
I W 1 

2 1 12 


1 3 
4 2 


1 12 


86 44 101 













Species A B C D E F 

Leucospora multifida I W 9 

(Michx.) Nutt. 

Linaria canadensis (L.) 2 2 

Dun . 

Ludwig ia mi crocarpa IN 11 1 


Ludwigia octovalvis (Jacq.) I N 116 21498 3586 307 23 19 

Raven and 

L. erecta (L.) Hara 

Ludwigia peruviana (L. ) I N 103 34 


Lysiloma ? 

Lythrum lanceolatum Ell. 

Mecardoni a montevi dens is 
(Spreng.) Pennell 

Medicago lupul i na L. 

Melanthera angustifolia I N 11 

A. Rich 

Mel ilotus alba Desr. 

Mikania scandens (L.) IN 42 


Mollugo vertici 1 lata L. 

Myrica cerifera L. IN 181 1 1 

x Oxalis filipes Small I W 61 18 15 200 11 9 96 

Panicum bartowense 
Scri bn . & Merr . 

Panicum tenerum Beyer I N 

Pani cum? 

Par thenociss us qui nquefo 1 ia I N 

(L. ) PI anchon 

Pas palum conjugatum I W 

Bergi us 

Paspalum s etaceum Michx. 

Physalis angulata L. 

Pluchea purpurascens (Sw.) 

Pluchea rosea R.K. Godt' I W 2 

Polygonum 1 apathi fo 1 i urn L. X W 901 3 1 

Portul aca o leracea L. I W 6 7 3 

l Pteris vittata (L. ) Small X W 2 6 4 9 26 

Rhynchospora divergens Chapm. IN 26 

Rorippa teres (Michx.) 

Sabatia stellar is Pursh I N 4 5 

Salix caroliniana Michx. I N 1 2 1 1 5 1 

Sambucus simpsoni i Rehder I W 15 11 



















1. 8 

1 r . 



























A B C D E F G 

Schinus terebinthi fol i us 

Sesbania macrocarpa Muhl . 

Solanum douglasii Dun. 

Solidago stricta Ait. 

Sonchus asper (L.) Hill 

Spartina ? 

Spermacoce tetraquetra 
A . Rich 

Sporobol us poirettii (Roem. 
S Schult.) Hitch. 

Tetrazygia bicolor (Mill.) 
Cogn . 

Thelypteris sp. 1 

Thelypteris sp. 2 

Trismeria trifoliata (L.) 
Di el s 

X W 

I W 47 54 
I W 5 

I N 

X W 1 

I w 

I w 

I N 

14 14 

3 33 

2 1 3 1 

2 13 11 

I N 


2 18 30 25 

1 1 

3 3 3 

Typha sp. 








scabra Vahl . 





2 4 

Vitis rot undi folia Michx. 























3300 22594 4611 1303 370 396 339 

Probable greenhouse contaminant 

older vegetation was striking, and involved values 
which differed by an order of magnitude. This 
decrease in seed density was to be expected, 
based on values from agricultural studies and 
early stages of succession, plus values from var- 
ious kinds of mature forests. Nevertheless, few 
investigators have measured the viable seed con- 
tent of different ages of successional vegetation 
in the same area. Two such studies (Oosting & 
Humphreys 1940 and Livingston & Allessio 1968) 
were based on buried seed; the surface litter was 
cleared away before the soil was sampled. The 
data of Oosting & Humphreys (1940) show a general 
trend of decreasing numbers of seed with increas- 
ing age of vegetation, while the data of Livingston 
& Allessio (1968) show no clear successional trend. 

Guevara & Gomez-Pompa (1972) found two-to-seven 
times more viable seed in successional stands than 
in mature forest in Veracruz, Mexico. 

The burned 2.5-year-old vegetation yielded 
about five times more seedlings per unit area than 
the 2.5-year-old unburned vegetation of the same 
age, an observation which may have important im- 
plications for the use of fire as a management 
tool. The burn may have increased Ludwigia spp. 
seed production, either by releasing nutrients 
into the soil from burned biomass or by altering 
the Ludwigia growth form, resulting in the 
production of more flowers and viable seed; 
another possibility is that burning affected the 
seed and/or seedbed, resulting in greater 
germination . 



r /////u//// /\ 

I — 10 

ii-ioo 101-1,000 10,001-ioopoo 

SIZE CLASS (no individuals) 

FIGURE 1. Number of species (all sites and all samples combined) in each of five logarithmically 
scales size classes. 


The number of species per site 
much less variable than the numbe 
Values ranged from about 30 to 50 
site, with no clear trends in spe 
reflected by successional stages, 
older communities had far fewer i 
the younger vegetation, yet simil 
species, the diversity (richness 
components) of the seed flora was 
the older vegetation types than i 
successional stages. 

(Fig. 3) was 
r of individuals. 

species per 
cies richness 
Because the 
ndividuals than 
ar numbers of 
plus evenness 

much greater in 
n the early 

Because the relationship be 
species encountered and area s 
and because different investig 
wide variety of sample areas, 
compare the richness of the se 
studies. Our values, however, 
range of those reported in the 
erally being higher than those 
forests and somewhat lower tha 
ported from the tropics . For 

tween numbers of 
ampled is nonlinear, 
ators have used a 
it is difficult to 
ed floras among 
fall within the 
literature, gen- 
from northern 
n most values re- 
example, Olmsted 

and Curtis (1947) observed zero to four species 
in six samples of 0.37 m 2 each in forests in 
Maine: Kellman (1970, 1974a) recorded 16 species 
per m and 19 species per 2.5 m 2 in coniferous 
forests of British Columbia; and Oosting & 
Humphreys (1940) reported relatively rich 
germinated-seed floras of 29 to 47 species from 
samples of 0.11 m 2 each from ten sites in North 
Carclina. Liew (1973) reported a relatively 
depauperate seed flora of only 31 species from 
20 m 2 of mature forest in Sabah, but most values 
reported from the tropics have been substantially 
higher. Guevara & Gomez-Pompa (1972), for 
example, recorded 14 to 28 species from samples 
of 0.51 m 2 in Mexico; Keay (1970) observed 42 
species on 1.15 m^ in Nigeria; and Kellman 
(1974b) reported a total of 54 species from a 
composite sample (78 sites) with a combined sur- 
face area of 0.23 m 2 in Belize. Our total flora 
of 119 species from a single area is higher than 
any reported in the literature, and is due, in 
part, to our relatively large total sample area 

of 5.25 m< 

The individual site values of 30 to 

50 species per 0.75 m 2 each, however, are quite 





2 5YR 2.5 YR. 17 YR 35 YR MATURE MATURE 



FIGURE 2. Density of germinated seeds from soils of the seven study sites after one year. 






u- 25- 



CO 20 






of the se 



van study s 


2 5YR 


17 YR 

35 YR 



weediness, and site-uniqueness of the germinated-seed flora from soild of each 
ites. Each value is the total of three 0.25 m 2 samples. 


60 r 










I 2 3 4 5 6 


FIGURE 4. Dispersion of 119 species among the seven study sites. 

high, especially considering that most of our 
sites are flooded for up to six months per year, 
which would seem to result in a great loss of 
viability for many species. 

Figure 3 also indicates the portion of the 
germinated-seed flora from each site which was 
made up of weeds. The definition we used to 
classify species as weedy or non-weedy was very 
conservative; anything specifically described by 
Long & Lakela (1971) as weedy or characteristic 
of disturbed sites was considered to be a weed, 
and all other species were classified as non- 
weedy. Because such definitions are quite 
arbitrary the weed flora at each site is shown 
on Figure 3 as a wavy line, rather than a precise 
number. The conservativeness of our classifi- 
cation resulted in some of our most abundant and 
ubiquitous species (e.g., Ladwigia spp . ) being 
classified as non-weedy. Only 36 percent of the 
seed-flora was made up of weeds, according to the 
criteria of Long & Lakela (1971) . The number of 
weedy species per site ranged from 12 to 27, 
making up 33 to 64 percent of the species. The 
younger stages of succession tended to have more 
weedy species, and these accounted for a higher 
percentage of their floras: usually half or more. 

Exotic species are particularly important in 
south Florida which, because it is a peninsula 
jutting into the subtropics, is potentially 
susceptible to species invasions from the tropics. 
Exotic trees such as Schinus terebi nthi folius , 
Melaleuca quinquenerv ia , and Casuar ina spp. 
dominate the landscape in many areas in south 
Florida and constitute a management problem for 
Everglades Park. We therefore suspected that 
exotics might be particularly important in the 
abandoned farmlands of the Hole-in- the-Donut , but 
were surprised to find that only about 10 percent 
of the species which germinated in the flats, 

representing less than 5 percent of the individ- 
uals identified to species, were exotics (accord- 
ing to Long & Lakela 1971) . This is even lower 
than the percentage of exotics in other floras. 
For example, about 18 percent of the 1647 species 
listed for all of south Florida by Long & Lakela 
(1971) are designated as exotics and 20 percent 
of the 5523 species listed in Gray's Manual 
(Fernald 1950) are exotics. Thus, although 
exotics certainly do become readily established 
in south Florida and often become dominant, 
conspicuous features of the landscape, the seed 
flora of abandoned Everglades farmlands does not 
seem to contain an unusually high fraction of 
exotics, either in terms of numbers of species or 
numbers of individuals. Of the exotic woody 
species which are important in and around Ever- 
glades National Park, the only one which 
germinated in our flats (from the two oldest 
successional stands) was Schinus terebinth! fol ius , 
which is already a major component of the older 
successional vegetation in the Hole-in-the-Donut . 

As shown in Figure 4, the species were un- 
evenly distributed among the sites. Almost half 
of the species which germinated were found on 
only one of the seven sites and more than 80 per- 
cent of them were found on fewer than four sites. 
Propagules of many of these species are un- 
doubtedly more widespread than indicated by our 
small samples, but at least they do not appear 
to be so widespread and abundant that they are 
likely to cause insurmountable management prob- 
lems. Rather, management practices might better 
be directed toward those 22 species which were 
observed in soil samples from four or more of the 
seven sites. These are the most ubiquitous, 
readily dispersed species and many of them are 
likely to become important components of the 
successional flora. 


On most sites, less than one-fourth of the 
species observed to germinate were unique to that 
site, as shown in Figure 3. One striking excep- 
tion was the mature glade; nearly half of the 
species tabulated from that site were observed 
nowhere else, even though the number of species 
which germinated from the mature glade soil (49) 
was unusually large. Apparently, many species 
which grow in the mature glade produce propagules, 
but these are not widely dispersed. The mature- 
glade vegetation is of low stature (usually less 
than 1 m tall) , so wind-mediated dispersal would 
be less likely to be as effective as it might 
from a forest. Also, many of the glade species 
may have propagules which are water-dispersed. 
Because the mature glade is at the southern border 
of the Hole-in- the-Donut, and because the prevail- 
ing drainage is toward the south - southwest, it 
is like